Dan Shipper interviews Jesse Beyroutey, managing partner at IA Ventures, about using large language models for investment research. They discuss how they made their 2019 Nvidia trade at $33/share, how LLMs like Gemini 1.5 Pro can be used to find patterns in large datasets and analyze the stock market via its 1M token context window, comparing Gemini 1.5 Pro head-to-head with ChatGPT, refining search queries for better AI answers, and new opportunities enabled by LLMs. They conclude by making a live $1,000 stock trade, with an epilogue recorded a week later. # Channel Presentation Observations — "How Do You Use ChatGPT?" (Ep. 13) ## Explicit Self-Description - **Channel name stated in outro:** The host closes with, *"smash that like button and subscribe to how do you use chat GPT"* — the show's title doubles as its premise. - **Host self-positions the show as a journey into AI's future:** The outro describes it as *"a journey into the future with Dan shipper as the captain of the spaceship."* - **The host is named Dan Shipper**, confirmed both in the outro and when the guest greets him: *"Jess welcome to the show… thanks Dan."* - **Episodic, guest-based format implied:** The title includes *"Ep. 13 with Jesse Beyroutey,"* and the host references past episodes, e.g., a prior appearance by *"Nathan leben who's actually on this show and did one of the most popular episodes we've ever had."* ## Demonstrated Scope: Practical LLM Experimentation - **Core content pattern: hands-on AI tool testing.** The episode's stated goal is to put Gemini 1.5 Pro *"through its paces and try to see what we can do with it."* - **Real workflows shown on screen:** stock screening via TradingView, prompt iteration in Gemini and ChatGPT, uploading ~100,000 tokens of earnings transcripts, and side-by-side model comparisons (*"I'm going to take this exact thing and put… into chbt and see if… let's do it"*). - **Comparative, multi-tool perspective:** The host and guest repeatedly evaluate Gemini vs. ChatGPT vs. Perplexity on the same prompts, e.g., *"we should do a side by side."* ## Demonstrated Scope: AI-Focused Interview Format - **Guests are domain experts presented with credentials:** Jesse is introduced as *"a partner at I Ventures… IA is a $600 million Venture fund and you've done seed investments in public companies like wise and digital ocean."* - **Personal, friendly rapport as presentation style:** The host describes the guest as *"one of my closest friends"* and frames the episode as *"trying to bring more Jesse into the world."* ## Intended Audience - **AI-curious practitioners, not domain specialists:** Content assumes familiarity with LLMs (*"this is not CBT this is Gemini 1.5 Pro which is this new model"*) while explaining investing basics. - **Casual, humorous register targeting a general tech audience:** The pair joke about WallStreetBets (*"treat this as worse than Wall Street bets"*), call the final trade a *"YOLO trade,"* and joke about hold music as an investment. - **Repeated disclaimers position a general audience needing warnings:** *"you shouldn't do this like you shouldn't day trade this is not a good idea… this is not investment advice use you probably buy index funds."* ## Observable Presentation Patterns (Not Claimed) - **Structured episode arc with a concrete deliverable:** A time-boxed exercise (*"we've got 90 minutes the clock is ticking"*) culminating in a real $1,000 trade placed live (*"swipe up to submit order received so there we go"*). - **Follow-up segments added post-episode:** The transcript includes a later segment where positions are exited and rolled: *"we actually made some money already on the 23 and Mi buy so I want to… exit that position."* - **Cross-promotion of adjacent podcasts:** A mid-roll plug for *"Cognitive Revolution"* hosted by a former guest, with specific named interviewees (*"Logan Kilpatrick of open AI… eugia cudia of replica"*). - **Emphasis on prompting techniques as teachable content:** The guest articulates methods like roleplay prompting (*"assume that you are such and such famous character"*) and iterative refinement (*"using llms is a task of continuously refining your question"*). ## Boundary Notes - No curated metadata was supplied; all observations derive from title and transcript. - Observations here reflect this single episode's presentation; channel-wide conclusions (e.g., consistent format, typical guests) would require further evidence beyond what this material demonstrates. # Delivery Observation: "Can Gemini 1.5 Pro Beat Our Best Stock Trade? - Ep. 13 with Jesse Beyroutey" ## Overall Format - **Conversational two-person experiment show**: host (Dan) and guest (Jesse) collaborate live on a task — using Gemini to pick a stock — rather than a traditional interview. - The episode includes **segments**: a cold open showing the trade result, a mid-roll ad read for another podcast, the main experiment, a follow-up segment recording a later trade, and an exaggerated outro. ## Tone - **Casual, enthusiastic, self-deprecating humor** runs throughout: > "we're like just two two dumb people on Wall Street BS trying to make some some kind of new idea here" > "a a stop clock is right twice a day or whatever you know" > "I feel horrible about it... me too" - **Warm personal rapport**: the host opens with an unusually effusive introduction: > "you're both uh I think one of the smartest people I've ever met... I'm so lucky to have you in my life" The guest responds in kind: > "completely overwhelming almost make me cry level of introduction" - **Playful irreverence about finance**: repeated jokes invoking WallStreetBets, e.g. > "treat this as worse than Wall Street debts" and > "this is what they call a YOLO trade" ## Recurring Phrases & Verbal Habits - **Heavy filler and restarts**, especially from the host: "uh," "um," "like," and repeated words: > "go go go gol internet" > "I I I I am" > "we've we've gotta we've gota make it we've got to make a bet" - **"I love that" / "I love it"** as the host's standard approval marker — appears at least six times ("I love that," "I love it okay cool," "I love that I think that's a really interesting place to start"). - **"Okay cool" / "all right" / "yeah"** as frequent transition and acknowledgment tokens between speakers. - **"Totally"** as the host's agreement word: > "totally and now today being able to do it..." — recurring many times. ## Teaching / Demonstration Habits - **Think-aloud narration while operating tools**: both speakers narrate their screen actions step by step: > "so I'm going to paste it in and hit go see what happens" > "we're going to swipe up to submit order received so there we go" - **Reading model output aloud and reacting**, then critiquing: > "thank you Gemini I appreciate that but you're saying there's a chance don't give up all hope just yet" - **Explicit technique-sharing about LLM prompting** is a stable pattern: - Roleplay prompting: > "assume that you are such and such famous character what would you have to say about X" - Forcing decisions: > "my immediate instinct is uh to ask it to just make a decision" - Iterative refinement: > "I find that using llms uh is a a task of continuously refining your question" - War-of-words prompting: > "you can uh give the llm basically War aibum text... stream of Consciousness and it can just do something with it" - **Skepticism and fact-checking of model claims**: > "is that true like we should probably check that... again something that we should check" ## Structural Communication Patterns - **Collaborative question-forming**: one speaker drafts a prompt aloud, the other refines it: > "how should I construct a screener to identify companies that might be artificial artificially depressed... does that seem like a good place to start?" - **Framework-building in public**: Jesse progressively articulates a multi-part investment thesis on the fly: > "it's like exogenous shock um sits in a key bottleneck of the value chain and and is able to scale" The host explicitly meta-comments on this: > "you have been refining the like key characteristics... I just want to like underscore that" - **Time pressure as a narrative device**: > "we've got about 15 minutes left" ... > "we've got under 10 minutes" ... > "Google we've got three minutes" — escalating urgency drives the finale. - **Frequent disclaimers** (stable pattern): > "this is not investment advice use you probably buy index funds" and > "do not do this at home absolutely do not do this at all" ## Comparative Tool Commentary - Recurring **side-by-side model comparisons** (Gemini vs. ChatGPT vs. Perplexity): > "should we do a side by" ... > "I'm going to take this exact thing and put Pace into chbt and see if... let's do it" - Repeated observations about **model hesitancy**: > "it's like it's a little bit hesitant for some of these queries to like really give us an answer" and later > "they're just getting a little bit frightened of of saying anything too direct" ## Uncertainty Notes - Speaker attribution between "Dan" and "Jesse" is **only explicit at a few points** (e.g., "Jess welcome to the show thanks Dan"); many quoted lines cannot be attributed with certainty to one speaker. - The outro ("I'm absolutely hopelessly in love with you") appears to be a scripted/parodic bit appended after the episode; whether it reflects the hosts' normal delivery is uncertain. # Observation Document: "Can Gemini 1.5 Pro Beat Our Best Stock Trade? — Ep. 13 with Jesse Beyroutey" ## Participants - **Dan Shipper** — host, "How do you use ChatGPT" - **Jesse Beyroutey** — partner at IA Ventures (described by Dan as a ~$600M venture fund; seed investments in Wise and Digital Ocean cited) ## Framing & Goal - The pair attempt to use **Gemini 1.5 Pro** (1M token context window, described as private beta) to replicate their best-ever trade: **Nvidia**, bought January 2019 at $33/share, later ~$800/share. - Original Nvidia thesis (read aloud): price down due to the trade war (temporary risk); Nvidia's proprietary GPU architecture and market position should grow with machine learning and gaming. - Jesse's qualification: they also made **one of their worst investments ever** around the same time; "it's not like we were absolute geniuses." - Dan commits **$1,000 in a Robinhood account** (his first); repeated disclaimers: "you shouldn't do this," not investment advice, only money they're willing to lose. ## Evolving Thesis (refined live by Jesse, highlighted by Dan) Three-part criteria for the target trade: 1. **Exogenous shock** — stock trading down for a clear, external, public reason (e.g., trade war for Nvidia; AT&T uncertainty for Meta). 2. **Sits in a key bottleneck** of the value chain (Nvidia/CUDA; Meta's at-scale advertising). 3. **Highly scalable business model** — could "grow 100x in a year" without hitting bottlenecks. ## Process Highlights - **Stock screener** (TradingView): $2B–$100B market cap, positive gross profit, down ~20% over six months, tech companies. - Gemini's first pass suggested **Bill Holdings, Klaviyo, Appian**; they struck Appian — Dan: low-code is "such a bad idea in an AI market." - Jesse's roleplay technique: prompt LLMs as famous figures (Jeff Bezos, Ben Thompson, Warren Buffett) whose perspective is informed and "a little bit contrarian." They run a Buffett simulation in Gemini; it declines to pick stocks ("impossible to make specific investment decision"), which they attribute partly to Google's safety posture. - Clustering exercise flagged **Gogo (inflight internet)**: hated brand but possible **hardware/distribution lock-in**. Key question per Jesse: **upgradeability**. Loaded ~100K tokens of 2022–2023 Gogo earnings transcripts into Gemini; management discussed **5G ATG network** and **Global Broadband (OneWeb LEO partnership with Hughes)**; upgrade requiring "only the installation of the ESA antenna and two wires." - Contradictions noted: Gemini claimed Gogo had strong stock performance and high multiples despite being filtered for low PE / depressed prices — Dan and Jesse flagged these claims as needing verification. - Jesse's turn: Gogo resembles their **worst** trade, not their best — distress was internal (5G chip design flaw delays, third-party subcontractor), not exogenous. - Pivot to **energy**: ChatGPT cited inflation data and energy sector weakness as exogenous factors; Gemini recommended midstream (Kinder Morgan, EPD), renewables (NextEra, First Solar), storage (Tesla, Fluence). First Solar repeatedly re-emerged, but Jesse noted energy companies lack the Nvidia/Meta-style scalability. - Neither model produced compelling less-known picks; Jesse: "we're gonna get a ton of stock pitches that are no good." - Time pressure (~15 minutes left) → pivot to **wrecked SPACs**. Considered OpenDoor, Virgin Galactic, 23andMe. ## The Bet - **23andMe (MEGH/23andMe)** at ~$0.58/share, ~$270M market cap. Rationale: bad scandal (data breach) as exogenous shock; genetic data potentially far more valuable with transformers; margin of safety from the battered price — **if it doesn't go bankrupt**, which Jesse flagged as the key unresolved question. - Dan explicitly: "I feel uncertain about this buy… we're doing it for the content." Jesse: "this is what they call a YOLO trade." - **Dan bought $1,000 of 23andMe** and both immediately expressed regret ("I feel horrible about it"). - Jesse's stated lesson: a good strategy doesn't mean an opportunity always fits; "you have to be really good at doing nothing"; wait for "fat pitches." ## Follow-Up Segment (post-episode) - **Exited 23andMe** at roughly breakeven-ish (down 4% on that position but up ~9% total in the account per Dan) — "we made money." - Rolled winnings into **~$1,087 of Google (Alphabet)**, arguing it fits the thesis: - **Exogenous shock**: Gemini PR problems (~$70B value loss claimed by Dan) plus disappointing earnings (weak ad revenue, unusually high capex) — Jesse: fears about LLM costs to search and about Google's internal culture were both "confirmed." - **Scalability**: Jesse argued Google designs its own TPUs/semiconductors and runs its own cloud, advantaging it for LLM-based search despite near-term margin pressure. - Dan's supporting analogy: sentiment resembled Microsoft ~10 years ago and Apple ~15 years ago, both of which recovered. - Qualification preserved: Jesse conceded Google is **up 50% year-over-year** and offers a **lower margin of safety** than the thesis ideally requires. ## Reflections on LLMs (both speakers) - **Massive context windows are only useful if content is accessible**; both repeatedly hit paywalls/broken internet ("the internet's so broken now"), attributing it partly to content owners blocking LLM indexing. - Both proposed a market for content: pay small amounts to push books/transcripts/data into LLM context ("buy a book for my LLM"). - **Real-time data** access would be highly advantaging. - Both observed LLMs are increasingly unwilling to make direct recommendations ("getting worse and worse over time"); Dan suggested opportunity for riskier, more decisive models; mused the "Golden Age" of LLM use may be now. - Performance note: Gemini described as slower than ChatGPT; Dan speculated requests queue behind Google's revenue-generating workloads (speculative, flagged as such). # Speaker Attribution & Provenance Record **Video:** Can Gemini 1.5 Pro Beat Our Best Stock Trade? - Ep. 13 with Jesse Beyroutey **Video ID:** 01a0f15b-31b1-79ef-a99e-f1efc92d3a95 ## Identified Speakers - **Dan Shipper** — **Creator / host**. Identified in closing promotional segment ("Dan shipper as the captain of the spaceship") and by Jesse's greeting ("Jess welcome to the show... thanks Dan"). Hosts the show, drives the experiment, executes the trades on a Robinhood account, and appears in the follow-up segment. - **Jesse Beyroutey** — **Guest**. Identified in the show title and introduced by Dan: "you are also a partner at IA [I] Ventures... a $600 million venture fund... seed investments in public companies like Wise and Digital Ocean." Dan notes Jesse is a close friend. ## Other Voice Sources (non-human) - **Gemini 1.5 Pro** — AI tool queried repeatedly; its outputs are quoted aloud (stock suggestions, clustering, 5G/GBB summary, etc.). Private invite beta; 1M token context. - **ChatGPT** — AI tool queried for comparison on several prompts. - **Perplexity** — AI tool queried once (on exogenous-shock stocks); result characterized as unhelpful. ## Promotional Segments (flagged) - **Mid-roll promo (~early in episode):** Promotion for **Nathan Laben's** podcast *"The Cognitive Revolution"*; named guests include Logan Kilpatrick (OpenAI), Eugenia Kuyda (Replica), Sual Doshi (Playground). - **Closing promotional segment:** Explicit "smash that like button and subscribe" call-to-action for the show "How Do You Use Chat GPT," ending with a declaration of affection toward Dan. Clearly flagged as promotional boilerplate. ## Referenced Sources & Third-Party Claims - **Nvidia trade (2019):** Bought at $33/share (Jan 2019), now $800/share — claim by hosts from personal experience; investment thesis quoted from a 2019 Notion doc ("biriri times shipper Investments"). - **TradingView** — stock screener used for filtering. - **Gogo earnings call transcripts (2022–2023)** — downloaded from Gogo's IR site; loaded into Gemini. - **Seeking Alpha** — source for 23andMe Q3 2024 earnings transcript; full article paywalled ("I can't read it because I don't have access"). - **SPAC Track** — website listing SPAC companies. - **News claims:** 23andMe earnings decline (revenue $44.7M, down 66.9%), delisting deficiency letter, possible company split (via web search); Gogo 5G launch delays attributed to a third-party subcontractor design error. - **Reddit/Google content deal** — mentioned by Dan as precedent for publisher–LLM licensing. ## Availability / Provenance Limits Noted in Transcript - **No real-time data in Gemini**; ChatGPT noted as having real-time access. - **Gemini cannot ingest YouTube links** directly; PDFs and audio/video files supported. - Analyst reports and some earnings transcripts were **inaccessible without subscriptions**. - Gemini **cannot provide links**; content had to be copy-pasted. - Speakers repeatedly caveat Gemini output as potentially fabricated ("it could be making it up"). - **Disclaimer:** Hosts explicitly state the trade is not investment advice and they only risked money they were willing to lose ($1,000; ~$1,087 on Google). - **Comment availability:** Not stated in transcript; no curated metadata supplied. when we bought Nvidia in January 2019 it was $33 a share today it is $800 a share that was the trade of Our Lives we want to find uh an investment that's just as good as our Nvidia investment using Gemini and I think we can do it basically what we have done is we have downloaded earnings call transcripts from 2022 and 2023 we're going to load them into uh Gemini and we'll see uh we'll ask some questions see how it does this is the stock that we came to at the end of a journey and I'll let you I'll let you see it as I do it we're going to swipe up to submit order received so there we go hey I want to just take a quick break from this episode to tell you about a podcast you should know about it's hosted by Nathan leben who's actually on this show and did one of the most popular episodes we've ever had about how to use AI as a co-pilot for the Mind Nathan hosts a podcast called the cognitive Revolution where he interviews Builders at the edge of AI and discusses the profound shifts that it will unlock over the next decade he talks to the researchers and entrepreneurs that are shaping This Moment Like Logan Kilpatrick of open AI who I interviewed for the show as well eugia cudia of replica and sual Doshi of playground to listen to the show search for cognitive revolution in your podcast player all right now let's get back to how do you use chat gbt Jess welcome to the show thanks Dan uh I am just delighted uh to do this with you um for people who don't know uh you are one of my closest friends um and you are also a partner at I Ventures uh IA is a $600 million uh Venture fund and you've done seed investments in public companies like wise and digital ocean um and you're both uh I think one of the smartest people I've ever met um you really like have this way of digging down into an issue and like kind of figuring out the Crux of it um and you're just like a very very good friend and very good person and I like just I'm so lucky to have you in my life and I'm so happy to be able to do this with you um yeah thanks for coming on I didn't know that I was going to get this uh like completely overwhelming almost make me cry level of introduction uh but thank you uh I also want to say thanks for getting me out of my natural ostrich-like tendency to not appear in anything or be on the internet um you always managed to stretch me uh in ways that have ultimately been so additive and productive in my life um you know the last time was helping me get one of my first things out onto the internet that people regularly reference uh back to me like oh I saw you on every uh so I'm really excited about this too great I just you know I get to hang out with you and hear all your smart thoughts all the time and I'm just like trying to bring more Jesse into the world so this is our latest our latest attempt um and you are coming on to talk about Gemini Pro 1.5 the new model from Google that has a million uh a million token context window which is like sort of wild and our goal is to put it through its paces and try to try to see what we can do with it um and the the concept that we came up with um uh is I think really uh really it's going to be really fun um and I'm just sharing my screen uh because uh to set it up what you realized is it would be really cool to do a good trade with Gemini and see if we can get it to like identify a stock to pick and it like hearkens back to this like I think probably the best trade at least I've ever done I don't know about you you're a professional investor it's mine too yeah maybe a second so in in in 2019 um we created a little notion doc called biriri times shipper Investments um because we had we wanted to take a bet on a specific company and the company that we wanted to bet on was Nvidia um and when we bought Nvidia in January 2019 it was $33 a share today it is $800 a share um and we wrote down our our our investment thesis we said you know the price is down because of the trade War but that's only a temporary risk um and that Nvidia is in a position of power because of its proprietary GP because of its proprietary GPU architecture and its position in the market should only grow as machine learning and gaming increases the need for high powerered CPUs and computers so the price dip should be temporary so so that was like that was the trade of Our Lives it was it was quite a moment I mean we we fancied ourselves as distressed asset investors in you know our little long short public market fund uh I think important to note that we also made one of our worst Investments ever at about the same time so uh you know it's not like we were absolute Geniuses uh and we didn't have llms at the time so you know maybe today'll be different yeah so so the concept for the show is we want to find uh an investment that's just as good as our Nvidia investment using Gemini and I think we can do it or I hope we can do it yeah uh I'm really excited about this in particular because um I love using llms uh for roleplaying type exercises as almost like a conversational participant along with me so my favorite sort of prompts for chat gbt and and now for Gemini are uh assume that you are such and such famous character what would you have to say about X and that elicits a different kind of perspective from the llm than just asking it a question flat out uh because it's it's trying to take advantage of whatever knowledge that it had about that person and things that co-occur with that person's name on the Internet totally and now today being able to do it with like an insane context window means that we can just load in like all of the context around a person and see if you know Gemini can take the the perspective of that person as if they're like an analyst on our little hedge fund team I love that um I'm curious like what are the names of people that you've used um and found to work well so far uh really depends on the setting yeah um but I mean I just generally always try to ask like what would the CEO of a famous company in that industry um so if I'm you know let's say I'm analyzing something about aggregators uh I might ask uh you know what would Jeff Bezos have to say about this uh or what would Ben Thompson have to say about this and it's kind of the intersection of someone who clearly would have an informed opinion about the thing uh probably has talked about it before has a lot of information about about them on the Internet already and therefore there's like some meaningful thing that the L might be able to draw from and um and then also that whose perspective is generally a little bit contrarian or comes at things from a different angle than um you know what the average of the internet might have that makes sense that's that's really cool yeah I love that um so I want to get us I want to get us going into our into our investment finding exercise because we we've got 90 minutes the clock is ticking um so set up for us um how we're thinking about like doing the initial selection of companies and I think we're basing it on how we thought about Nvidia and we're going to try to see if we can do a similar kind of filter for this this round of investment right exactly so um back when when we were doing our little hedge fund in 2019 uh our idea was we wanted to find companies that had been trading down suddenly recently for some reason that we could look at and think oh this distress situation is temporary and and so from our perspective as a more like long-term investor it gives us uh uh the advantage of a margin of safety oh if it's trading down by 20% 30% or whatever then we can be a little bit wrong about things and uh and still not lose too much money as happened with our uh worst case investment that we made around that time where it took us a little while but we eventually recovered our our principle on it and we're able to live to to Die Another Day yeah totally and I will say like yes we we went into this with this very long-term thesis like oh it's only down temporarily but then I sold I panic sold it during the pandemic so like I I I look stupid but like I still feel good about identifying it originally but I'm I'm gonna hold this one this this trade is I'm gonna I'm Gonna Keep around I think we should also say for anyone who's listening you shouldn't do this like you shouldn't day trade this is not a good idea uh and we're doing it only with an amount of money that we're willing to lose um so this is not investment advice use you probably buy index funds unless you really know what you're doing but anyway um yeah treat this as worse than Wall Street debts because it doesn't even have the common thread and more on that later because we might even post our bet to Wall Street bets we are going we are going to probably post our bet to W stre bet we can but um one thing that I forgot to say is we are going to put actual money down so I created a Robin Hood account which is my the first time I've ever had a Robin Hood account I put a$ thousand dollars in and hopefully by the end of the episode we will find a company to put $1 thousand dollars in and we'll be able to track how it does cool um so yeah why don't you why don't you I think you're I think you've set up the filter of what we're looking for appropriately and I know you're you've been looking through companies to to kind of start our research with do you want to share your screen and and show us what you've been looking at yeah let's let's open it up here so uh I just popped open this website trading View and um it gives you a way of screening all of the available stocks against a bunch of filters and criteria and when I thought about the the context of our Nvidia investment um we were looking at companies that weren't massive yet but had the potential to be the you know 10 bagger of the future uh and so for me that was expressed in the form of it's a company that's between a couple billion dollars and 100 billion in market cap so it could become the next trillion dollar company uh but it's starting from Small Beginnings I think Nvidia was like something in the 50s 100 billion range uh when we first invested and now it's a$2 trillion dollar company today depending on the trading uh so I was looking for for companies like that uh that were uh in positive gross profit which we might play around with but just you know want companies that actually make a little bit of money they're not gonna die yeah yeah like that you know we we're looking to avoid our total Bozo risk as much as we possibly can and uh the way that we're trying to do that is just like okay the company has to figure out how how to make money somehow right totally and then um aside from that I was just looking for uh like really there's so many like screens that you can do but um I really just wanted to find companies that had a six-month performance that was like they're down 20% or something like that but then you know on a fiveyear basis in in this column uh you know are are are also down but um you know we might come up with a reason for why that was the basic screen and we only looking at technology companies because that was what You Know Dan and I thought maybe we'll have some something posi whatsoever but I think we should also try to do companies that we we know absolutely nothing about later right totally so I'm looking at this list and the first thing in my mind is like I guess we could just throw this into Gemini and ask it to filter based on our like Nvidia criteria yes but I'm I'm I just want to confirm that that that seems like a good idea to you I think anything with llms is a good idea because we'll see what we get out of so let's just do it so somehow you're going to have to get that that uh that stock screener screen over to me um so I throw it into Gemini yeah uh all right so what we've got here is you copy pasted your uh the output from from the trade View and I'm going to just head into um head into Gemini and so again this is this is not CBT this is Gemini 1.5 Pro which is this new model it's it's invite beta only private beta um and what we're going to do is we're going to take the um we're going to take the output uh that that you gave me and we're going to try to filter it by our criteria so I'm gonna let's say something like hey I'm trying to uh pick a stock I want to find a 10 bagger specifically I'm trying to replicate um a trade I did in 2019 where I bought Nvidia it was down temporarily because of the trade War um and now it's a$ two trillion um now it's a two trillion doll uh company um here's a list of potential starting points for trades um does anything pop out to you any uh any any revisions there that seem like a good Okay cool so I'm going to paste it in and hit go see what happens one of the things I've noticed about uh about Gemini is it's slower than chat so like it takes it can take a million things can take a million tokens but you have to wait like I don't know sometimes it takes a minute for it to actually return which is I mean it's because our request just got queed behind all the things that make Google money to run on their tpus or what ever so I get it okay so Google is uh Gemini is saying um good caveat it's important to remember that past performance is not indicative of future results and finding the next 10 bagger is extremely difficult while your Nvidia trade was successful replicating that exact scenario is unlikely thank you Gemini I appreciate that but you're saying there's a chance don't give up all hope just yet um however I can offer some insights on the you provided potential growth stocks bill D Bill Holdings Incorporated this company provides cloud-based software solutions for managing invoices and payments it has shown strong Revenue growth and has the potential to benefit from the increasing adoption of digital Payment Solutions um uh clavio this company offers an email marketing platform for e-commerce businesses it has a strong track record of growth and is well positioned in a growing Market appen this company provides a low code platform for building business applications the demand for low code Solutions is increasing which could benefit appen um important considerations high growth potential look for companies with strong growth valuation while while high grow how while high growth companies often have high valuations it's important to ensure the price is Justified okay so I think so it's it it identified Bill Holdings clavio and appion I think we should strike appion because I just feel like low code is just like such a bad idea in an AI Market um but I'm curious if you have a differing opinion and or if any of the other any of its other suggestions pop out to you is something we should explore yeah uh so what immediately stood out to me is that it's um it's it's already identified that uh high growth companies have uh potentially High valuations uh which is obviously true and here what we're specifically looking for as a starting point is a company that's artificially depressed right and so we want to figure out what reason there might be for the company to be artificially depressed and give us that that you know nice feeling that we can make our money back even if were totally wrong right right so uh I would actually even let's like try a different angle um the angle might be uh instead of um asking about those specific stocks let's ask it for uh how to construct a screener to identify companies that might be artificially depressed in their prices because maybe my screener idea is actually not even the right one okay um and should we say should we use the the trade view um and say like in the trade view or um can't hurt I mean it might aare of what that is but but also even if it's not um you know stock screener type filters are pretty generic so I think it could just give us some ideas okay cool how should I construct a screener to identify companies that might be artificial artificially depressed in a stock screener does that seem like a good place to start okay and I will say like these are things that we could do with chbt um and hopefully once we narrow things down a little bit we will uh start using more of Gemini's like in like increased capacity to to analyze data all right so here are some criteria you can use to construct a stock screener price to earnings ratio look for companies with uh PE ratio significantly lower than their historical average price to book ratio similar to PB to PE a PB ratio significantly lower is significantly lower than its historical average would be good price to cash flow ratio debt to equity ratio free cash flow um it's talking about technical IND indicators like the relative strength index I don't really want to use technical indicators um uh so I don't know any of those any of those things spark something for you we could also think about um you know throwing in something like uh like Ben Graham's intelligent investor book and see uh see what how Ben Graham might evaluate um some of these stocks but uh but yeah uh tell me what you tell me what you think my my immediate instinct is uh to ask it to just make a decision uh so like every time I ever get some kind of summary from chat gbt or Gemini uh it's always like caveat heavily like you might do this or you might do that and I'm like okay great I'm just trying to do the the best version that I can with five uh parameters like what would you use okay so make okay make a decision do the best uh you can yeah okay cool you kind of want to download the intelligent investor just have it based on the information provided it is imposs to make specific investment decision choosing a stock so this is you know I think um it's you know Google's model is going to be probably pretty safe you know yeah yeah we're we're running up against some uh Google doesn't want to be liable for what we did with G type issues which I get it you know we're like just two two dumb people on Wall Street BS trying to make some some kind of new idea here but uh you know I think I think we should try to get it to to be a little bit more explicit we could say like pretend you're I don't know Warren Buffett or something yeah um and I've come to you to make a final decision on you want to have it decide on a trading strategy or on a stock I think we should go with trading strategy as a starting point trading strategy to start with and this will probably weight it towards Warren Buffett's like trading strategy yeah um but I think it'll um I think it'll get it to make a decision for sure does it work with links I don't think so okay but we can just copy paste anything that's in a link into the does it work with PDFs yet or no it works with PDFs okay got it sweet it also works with videos um so interesting so if you've got any videos um or audio also great uh so my instinct is I think it'd be really fun to have it um try to get like uh people who run major hedge funds to be our analysts I love that so uh yeah okay since he's been in the news a lot I'm trying to think of a good thing to feed in for Bill Amman okay let's let's have let's get Bill's opinion B find a good distressed stock right now um so I'll say Warren Buffett in Gemini says well I don't advocate for frequent trading if you insist on starting a strategy consider a value oriented approach look for undervalued companies using metrics like PE PB and and P overc PCF um look for companies with strong fundamentals and a durable competitive Advantage so these are all things that we've already kind of considered but I think the PE maybe a PE uh a low PE filter is an interesting one to consider yeah yeah so uh let's I'm just going to try fil so I broadened our industry here and um I'm going to try a uh price the earnings filter so basically Gemini our simulation of Warren Buffett asked us to filter our uh our list by uh price to earnings ratio uh you know below a certain number we picked 15 and so these are the companies that are that are in that bucket and these are the companies that um have a fairly low price earnings ratio might be depressed I have not filtered this by tech companies right um and maybe that's purposeful I just want to make sure you're you're good with that yeah I mean I think we we can do um uh like uh a broad screen and just have it have it tell us what what to do like a like an analyst would perfect um one that I'm feeling is there's not we don't have real-time data in Gemini so asking it about like what is um currently depressed might not work that well well we're going to use the stock screener to figure out what's currently depressed so the the way that I was thinking about doing that was um so they have if you search filter it by the word performance they have uh ways of screening by six Monon performance okay cool co co ET and so what I was thinking is okay we want something that's uh you know below minus 20 on the six month and um and then you know on a fiveyear basis maybe we want something that is um you know at or above zero yeah I think that's I like that because it it indicates strength but temporary dysfunction exactly myap told me that last week so that's where that came from um so Okay so we've got 29 matches um and we could look at them individually ourselves we could throw them into Gemini what's your what's your call yeah so I I think that now is a good time to uh to throw them into Gemini and ask ask it to look for uh points of commonality or something like that okay cool uh points of commonality tell me more about that well um as a starting point um I think that language models are amazing at clustering things so um if I if I were to ask it to um you know group them into some relevant groups that we can consider if you had to Cluster these stocks how would you do it and why all right so I'm just going to paste the plain text and see how that goes okay cool so okay so it says clustering stocks involve grouping them based on similarities here are some potential approaches in in the ration now clustering by sector clustering by Financial metrics clustering by technical indicators clustering by fundamental factors hybrid approach choosing the best approach the best approach for clustering stock depends on your specific goals and strategy for long-term investors focusing on fundamental strategies and financial metrics for short-term investors technical indicators and price Behavior might be more important so um I would actually ask it uh because obviously those things are true but but I would actually ask it to to Cluster them by things about the company and what it does okay um so that it's more possible for us to analyze it okay more um on the basis of just the the nature of the company and its position and its industry the nature of the company and its position in the in the industry so I'm just literally typing what you said and see what it does all right let's try it um and I'm noticing that it's like it's a little bit hesitant for some of these queries to like really give us an answer yeah I'm not sure that chat gbt would be so hesitant all right well let's let's side we should do a side by while this is loading okay so uh so Gemini said uh clustering Stocks by company nature and Industry position so based on the information provided here's a possible clustering of of the stocks based on the nature of the company and its position in the industry cluster one commodity based businesses companies like Archer Daniels darling ingredients civitas blah blah blah these companies are primarily involved in the extraction processing or distribution of Commodities like agricultural products this immediately makes me think of lithium um for Batteries y or anything and anything related to like chips or like Silicon or whatever seems like an interesting play but I bet those are already hot I don't think they're depressed right now right yeah we're going to have a hard time finding anything depressed that's around semiconductors because it's so on thesis right now that makes sense okay um these companies are typically price takers with limited control over the market price of their product their performance is heavily influenced by supply and demand Dynamics and global commodity prices that doesn't sound like our kind of invest think so um consumer facing businesses companies Davis Budget Group car I hate budget I will not invest in budget this is like like a moral stand yeah we we're taking a stand here we're only investing in a company that we actually like or at least one that I haven't heard of or at least one that I haven't had a personally horrible experience with um so that's that's funny because out of the whole list the one that stood out to me was go go go gol internet I was thinking about how it's potentially artificially depressed because uh it's both a a brand that people hate everyone has the same reaction that you just had to budget and also so at the same time there's like the Buzzy starlink company that's going to try to go after them and so I wonder how much that that impacts them that's an interesting thing to to look at I love that okay cool uh Den mode which I have no idea what that is um we could look it up Urban and urban one um so those are the other uh consumer facing businesses aside from goog go then there's technology oriented businesses so this is our bre and butter right here um perryon Network intest Corporation Universal Security Instruments and Electronics these companies developer provide Technology based products and services including software um and then we've got a couple more uh Dita uh in healthcare Dita CrossCountry Healthcare amen Healthcare Services I know you're you've got some investments in healthcare I don't know if any of these pop out to you yeah I mean I think a lot of people really dislike Dita so that that's another one um in fact actually what I was thinking of asking Gemini is um which of these businesses are most likely to be uh uh trading poorly because consumers dislike the brand H okay so which of these businesses are most likely to be trading poorly because consumers dislike the brand um because I think what you want to do is just strike them from our list maybe or may that's actually worth yeah tell me about that though because like if consumers dislike the brand theoretically that leaves a lot of room for a competitor to come along that would like kill the company right yes and no so uh if consumers dislike a brand but it has some crazy distribution lock in uh uh so you know go go being a great example maybe um go go is is the default and has Hardware lock in on so so many planes that it's not going anywhere anytime soon even though consumers hate it and uh and so that maybe is an interesting opportunity I get it okay so and it it said that so from the provided list the businesses most likely to be trading poorly are go go go go provides inflight internet and entertainment services Urban one Urban one is a media company targeting African-American and urban audiences and Davis Budget Group while car rentals are generally more susceptible to economic fluctuation in travel demand negative experien as a customer service vehicle could damage brand perception um all right I I I'm I'm getting a feeling here which is uh I just think that goo is going to be interesting for us to analyze so I want I want I want to dig straight in start talking about goo okay how do we do that yeah so uh I think the key question for goo is um whether or not their technology can be upgraded so uh it's clear that they have a distribution Advantage already right because they're already in all the airplanes they're in all the airplanes and uh they've also integrated with all of the uh apps and services that the Airlines use to provision the internet and and connect your payments and all that sort of stuff so they're kind of the one to beat and um people hate the service because the service quality is low right so the question is upgradeability and and so I think it'd be interesting to to like start talking through with with uh Gemini how go go might might go about upgrading I love that I think that's a really interesting place to start why didn't you think about first like Arline switching costs or like how hard it is for the airlines to rip them out and move move to someone else uh my my immediate thought was that if go go is incumbent uh it's the one that they already have a deal with uh it's the one that's already in installed and if there's some clear upgrade path for them um that's going to be the natural first step for the airlines uh before bidding out completely replacing all the hardware and so if it's possible for there to be a a lowcost upgrade path I think that they they have that default stat manage okay that makes sense Okay cool so should I just ask that yeah Okay cool so um H how hard would it be for uh go go to produce Hardware um that would allow it to upgrade its current services in Planes um to what what's the to be like starlink technology uh yeah to to improve uh service quality uh using the the latest in connectivity like starlink okay like starlink all right let's see and then I think this is also a good opportunity to start integrating some uh of Gemini's giant processing cap ility do and um my immediate thought is let's just like get the earnings transcripts yes and see how much they're talking about this okay great how do I do that uh just go to their website and grab the earnings transcripts I think is the best way but we could also ask Gemini if there's a faster way to load them all in and see what says no I just I've never gotten anyone's earnings trans have to show me I mean literally just like uh I mean the the the dumb and simple ways to find their IR site and go to their quarterly results okay cool inv relations here we go we got it earnings transcript um uh do we want annual reports earnings reports yeah earnings and then um we want let's get the earnings call yeah yeah yeah exactly so the the the transcript usually has all of the conversation um already transcri transcribed okay but we could also ask Gemini to to to listen to the audio and see how they sound well what I want to do what I'm really interested in is um is whether and I can like detect the um like the the voice oops you guys um can detect like uh the voice voice tone and stuff like that to see like oh are they like really um nervous or whatever you know like how are they feeling yeah and I'm gonna have to do some magic to see if I can actually download this file um but yeah that doesn't seem like it's going to be super easy but I might be able to find it somewhere else so okay so do I do Ann uh go go announces third quarter results go go earnings earnings transcript that's transcript yeah yeah uh but also by the way I just um I just found uh it seems like there's a company that publishes the conference calls onto YouTube oh sweet um so I just tried searching YouTube uh for uh go go earnings call right and I I can download a YouTube uh a YouTube downloader yeah um oh Gemini can't just like go straight in and process the YouTube link I see got it oh actually never mind uh it it only had a couple of random ones I thought it was every single one okay all right different strategy pivot pivot we've got some we've got a transcript so I have that now we can start with that or if you want to keep looking for more stuff um there's we can look at like Q well I'm thinking let's actually try to load in like a bunch of the transcripts at once you want to do the 10K from 2022 sure but but I think the transcripts uh have some more like um uh there's Q&A in there and what I'm really wondering is how much that uh things like replacement cycle and starlink as a competitor and whatnot are coming up and what Management's sort reactions to those are this is so good okay so I have all the earnings transcrip I have all the earnings transcripts from 2022 and 2023 is that enough to start let's do that okay I'll start with that yeah um cool so I'm downloading save I realized that uh our potential bet on go go here has the characteristics of our worst investment more than the characteristics of our best investment interesting our worst inv investment came from the reason for a temporary dislocation being that well uh this company was being sued and uh there were a bunch of reasons why all of a sudden people were were thinking negatively of it because of that sort of recency of announcement um and it it wasn't benefiting from a wave of uh positive momentum for the adoption of a new technology that it was already well positioned for and so I'm wondering as we dig into it a little bit more as to whether we can find reasons that Gog go is actually going to be a huge beneficiary of that new technology wave or if it's actually uh disliked for good reason right right I love that well to that point so we have a response from Gemini about wow um about uh how how hard or easy it would be to upgrade its hardware and so Gemini says for goo to produce Hardware that would allow it to upgrade its current services and planes to improve service quality several factors need to be considered compatibility goa's current systems operate on groundbased cellular networks while starlink utilizes satellite constellation integrating these two technologies would require significant engineering effort Hardware development developing new hardware that can communicate with Starling satellites and integrate with existing aircra systems would be complex and expensive regulatory approval so I mean we have to take all this with the grain of salt it could be making it up of course for all we know go go has this well let's let's go to the transcrip we should Google or go to let's go to transcript okay so basically what we have done is we have downloaded all of goo's earnings call transcripts from 2022 and 2023 we're going to load them into uh Gemini and we'll see uh we'll ask some questions see how it does so um we're going to take this and we're going to throw it in here and um we'll see it's all being uploaded what what would be your first question of of these transcripts I would ask uh to what extent is management already talking about upgrading nice okay um upgrading to new technology how's that let's do it great so I'm just saying to what extent is management already talking about upgrading to new technology and uh I think we just put in a about 100,000 tokens so this is already more than chat can handle um and it's about where uh open uh latest model which is not yet available generally in chbt it's about what it can handle it's about 128,000 tokens GPT turbo so we're pushing the limits of existing technology already let's do it yeah again we need the uh the waiting music we need the uh the already baked um um the already baked response to take out of the oven okay so management has been actively discussing upgrading to new technology specifically in reference to their 5G atg Network and Global Broadband initiative here's a summary of their comments from the q1 2022 earnings call 5G atg Network go go on track for commercial launch in the second half 2022 they have passed all development risks and are focused on a 5G chipset tracking and building Towers goo is already selling shipping and installing 5G equipment that can operate on the 4G network until the 5G box is available customers who pre-provision with with this equipment can easily upgrade to 5G later by simply swapping the Box goo seeing strong demand for 5G with orders from end customers dealers and commitments from oems Global Broadband gbb initiative goo is partnering with Hugh Network systems to develop a small electronically steered antenna to access the one web low earth orbit Leo satellite constellation this will enable goo to offer Global Broadband con connectivity to business Aviation significantly expanding their adjustable Market gbb is expected to launch commercially in the second half of 2024 goo is seeing overwhelmingly positive responses to gbb from customers dealers and oems for existing advaned customers the upgrade to gbb will be simple requiring only the installation of the esa antenna and running two wires inside the aircraft so that's actually a really interesting point for what you were just saying earlier earlier that it will be pretty easy for them it looks like from this one search uh to upgrade their uh upgrade their air aircraft to uh to satellite fantastic well that that's that's looking pretty positive I could imagine uh you know people not following that specifically um now um let's ask uh to to what extent would um that upcoming change already be priced in to their stock and how would we figure that out okay that's that's a great idea uh to what extent would that upcoming change already be priced into their stock and how would we figure that out um I think it' be worth also inquiring as to like why is the stock down right now yeah like because we're assuming that it's probably because of starlink but maybe it's something else maybe could Global avation not doing well you know to totally other reasons all right it is difficult to say definitively to what extent the upcoming technology changes are already priced into goo's stock the stock price reflects a variety of factors including Current financial performance blah blah blah however there are some indicators that the market is anticipating the positive impacts of goo's new technology strong stock performance goo stock price has performed well in recent years significantly outperforming the broader Market is that true like we should probably check that um this suggests that investors are optimistic about the company's future growth prospects positive analyst sentiment the majority of analysts covering goo have a buy or strong buy rating again something that we should check um High trading multiples goo stock trades at relatively High multiples compared to its peers this suggests that investor willing to pay premium given that we filtered by low uh low PE ratio companies I'm surprised that it's um has a high trading multiple I agree um so here are some approaches to gauge the Market's expectations analyze analyst reports compare goog go to its peer and peers and model the impact of new technologies so what do you think we should do you want to start with analyze analyst reports or is there a follow-up question you want to ask yeah well I'm I'm particularly trying to think about ways that we can use the advantage of a giant context window um and analyst reports would have been great if we had an easy access to them but I I don't know if we're gonna be able to easily download like okay all them you don't have any like subscription to like don't don't do public stock so I got got as much as the average reditor does okay um so I I don't think that that's gonna be a useful Direction um where I'm at right now is uh I'm thinking about how um we're trying to establish why go go is down yeah um and uh that's because we started with a stock screener and just looked at things that were down yeah uh but if I think about our original Theory generation strategy it was finding things that were down for a clear reason that we knew and the reason being something that was sort of public and in the news um so now I'm starting to think that maybe a different approach would be better where we're actually starting from uh more like recent news events or things going on in global macro that might cause uh something to artificially trade down that we can specifically find I love that okay um we we can totally do that uh I think we would probably use CBT for that all right let's do it let's Swit tools but unless you had a sense for like how you wanted to go about no I didn't have any sense more than I said yeah let's try it shot and we'll see uh we'll see I don't want to lose the like the go go thread there's something in it there's something in it uh I'm just realizing that in order for us to establish why it's down uh and what is or isn't priced in um we're trying to hunt for a specific reason for it being down yeah and um I think we could potentially establish that but it'd be hard to feel confident about it I I I just did a little search and one thing that's coming up already is ah go go delays launch of inflight 5G Network for second time in 12 months and so if we had if we had to guess connecting the dots I'd say that's probably part of it yeah is um uh and they attributed the delay to a design error made by a third party subcontractor of its solution provider inflight internet provider goo business Aviation is delaying the launch of its 5G Network originally slated for a fourth quarter 2022 rollout the service was delayed to Mid 2023 to a computer chip testing holdup now goo expects a design flaw in its chips to push the launch back further to Mid 2024 so I think like one of our one of the things in the calculus is do we think that they'll end up executing on this is the are the endless delays signs of incompetence or um do we want to bet um yeah yeah yeah I I I definitely see that and um I guess the the thing that feels important to me here is that delays are a uh an internal thing uh and what we're really looking for is a stock trading down do an exogenous shock Okay cool so great example you know inid it was the trade War yeah uh you know more recently Meadow is an amazing trade because it was artificially down due to at and uncertainty around it you're totally right so I I'm I'm really hoping we can find some exogenous factors right now that might be affecting affecting stocks okay what are some stocks that are trading down because of exogenous factors and should we maybe say tech stocks or just any stock try any for now I'm going to do this in chat gbt and I'm also going to do it in perplexity all right because I don't know my perplexity might work actually um uh so let's see perplexity first okay proplex is not not answering this question well um it's basically telling me why stocks might trade down rather than using stock prices um and uh some stocks are trading down due to EXO chbt says some stocks that are trading down due to exogenous factors in 2024 include those affect by inflation data and the energy sector stocks inflation data impact stocks in general have experienced downward pressure due to higher than expected inflation data the CPI data for January showed a 3.1% year-over-year increase so um so inflation is is one big exog factor and then energy energy stocks after leading the market in 2022 took a breather in 2023 losing 6.7% by mid December compared to a nearly 20% gain so what is this making you think uh my immediate instinct is um well if if energy stocks are artificially down because of a bunch of of exogenous factors the second the part of a thesis is uh that also we want companies that are going to benefit from a positive wave of adoption uh due to uh a technology that they've built being adopted rapidly so energy strikes me as an interesting one to dig into because uh there's a bunch of Technology shifts on the horizon for energy and like government investing in it a lot at the same time that there's maybe some reasons why the the energy stocks are down so are you thinking like um like big energy stocks like Exxon are you thinking like trying to find some like Solar Company or some wind company or nuclear or something like that um well I I think it's important that we find a company that sits on a key bottleneck that was the Insight behind Nvidia and uh meta as two example uh trades was that uh Nvidia had this huge advantage in uh Cuda in its particular case and you know meta had this advantage in being the highest at scale player for for advertising for a certain type and you know in both cases they um they stood to take advantage of their position in that value chain so um I think we should try to find something that that might be similar okay and um how would you how would you suggest we go about that um should we just should we ask like what energy companies control like a key a key bottleneck um yeah let's do that okay and ideally want to frame it in reference to and our position to ride a new technology wave yeah um what energy companies um control a key bottleneck in the value chain phrase it like that um in the value chain um and are positioned to ride a new technology wave over the next five years let's say is that does that sound good okay and Chach BT at the bed the best analysts do that too can't can't help ourselves sometime all right let's see uh we're we're waiting for chat gbt generation but I think while chat gbt is is thinking we could um we could just go over to back over to Gemini and throw that in as well and and compare yeah so I'm GNA ask Gemini what energy companies control key bottleneck and the value chain and our position to do ride a new technology wave over the next five years in the energy gbt is saying in the energy sector companies that control key bottlenecks in the value chain and are positioned to ride a new technology over the next 5 years include those involved in hydrogen production and Fuel Cell technology as well as those focus on building resilient Supply chains for energy transition Technologies so hydrogen production and Fuel Cell technology hydrogen is emerging as a key player in the energy transition with applications from Renewable Energy Systems to decarbonizing Transportation companies involved in hydrogen production fuel fuel fuel cell technology and related infrastructure are well positioned to capitalize on this hydrogen power vehicles and Fuel Cell uh electronic vehicles are expected to play a complimentary role okay building resilient Supply chains as the energy transition accelerates supply chain resilience is becoming increasingly important companies that can navigate and mitigate risks related to volume shortages price volatility geographical sourcing dependencies long lead times and quality issues in the supply chain are will have a competitive advantage and then green business opportunities companies that can scale green technologies quickly and efficiently will have an advantage um what's what's popping out to you well we need to ask it for public stocks so let's let's ask you for examples of public stocks that might benefit in each of those categories okay give us some public stocks that fit each of these categories we'll see what Gemini said um so Gemini by comparison said here are some energy companies that control key bottlenecks in the value chain are position to ride a new technology over the next five years one is Midstream companies bottleneck midst companies control the transportation and storage of oil and gas this is actually pretty cool um this gives them significant leverage in the value chain as it can influence the flow and pricing of these Commodities new technology waves Midstream companies are investing in new technology to improve the efficiency and sustainability of their operation this includes pipeline Leak Detection automated pipeline operations and carbon capture and storage examples are Kinder Morgan and Enterprise Products Partners like I like that it's it's giving examples yeah it already went to the examples clustered them in a useful way yeah it's pretty cool um renewable energy developers bottleneck renewable energy developers control the Development and Construction of renewable energy projects this gives them a key role in the transition to low carbonate Energy new technology wave renewable energy developers are constantly innovating to improve the efficiency and cost effectiveness of renewable energy Technologies um so solar panels wind turbines energy storage and it's suggesting next era energy the world's largest producer of wind and solar energy and then First Solar a leading manufacturer of thin film solar panners fslr is constantly innovating to improve the efficiency and affordability of its products energy storage companies bottleneck energy storage is a critical bottom neck and the transition to a renewable future as the grid becomes increasingly reliant on intermittent renewable energy sources energy storage is needed to balance the supply and demand new technology wave lithium ion batteries flow batteries and hydrogen storage and it's recommending Tesla and fluence energy um yeah what are you thinking uh well I'm thinking let's let's find out if any of those are are artificially down yeah exactly okay cool um I wonder what the um oh let's let's compare to chbt so chbt says plug and power for hydrogen production it says plug and plug power ink ballor Power Systems and Bloom Energy and then for building resilience Supply chains it says alour Corporation Freeport McMoRan incorporation riotinto Group which who knows um and then green business it recommends beyond meat Tesla and First Solar so actually the green business is the is the one sort of overlap between Gemini and um and chbt what I'd like to do is have it um uh I'm gonna here's uh I'm GNA have Gemini just output a clean list for us here's a list of other recommendations please please print them out um please print these and your recommendations as a bulleted uh list with just the ticker symbol just so it makes it easy to add to our stock screener beautiful so um and then I'm going to split my screen and what was the stock screener again uh trading view trading view okay I'm not ready um and then how do you in trading view look at a uh specific set of ticker symbols filters uh I think you can I think I just type in a symbol yeah I I don't think there's like a faster way of doing it okay um we've got our clean list of recommendations um yeah how should we start just go down the list yeah I mean I'm trying to figure out a way to upload them okay oh upload them or like you know just dump a list into a product but I can't find one okay if there's any way okay well we'll just go one by one unless there's I'm happy to copy paste them individually because there's not that many if we can get them all up on a screen somewhere but I don't know what product would do that me neither okay cool um well we'll just we'll just do it this way so let's go one by one so we're going to start with Ki and um we're g to just kind of do a little spot check of Ki uh looks like it is down um we've got a$ 38 billion market cap uh where is the PE Ratio is 16 so it's like just above our threshold um dividend yield is 6.55% um yeah anything uh anything sticking out to you here or should we move on to the next one uh yeah it just doesn't look like it's uh it's gotten hit by any kind of recent exogenous event and that's really like a key factor that we're looking for so yeah it's down 10% over five years and it's down only 0. 29% okay so yeah over the last six months so I think we're I think we're good we're good on that one right um so I'm G to just pop this out I think it's not really working that well uh I don't know how to do that separate Okay cool so next one is epd so epd is um Enterprise Products Partners LP um this is also up six 5% over the last six months so that's not really that's not really that good um let's see NE next era energy Inc um down 15% over the last uh 6 months up 20% over the last 5 years so we've got strong company potentially over the last five years but definitely hit by an exogenous shock um it's next era energy it's a utility company it engages in the provision of renewable energy it operates through the following segments FPL and Neer I have no idea what that means um the FPL segment involves the energy transmission distribution and sale of electric energy in Florida the near segment produces electricity electricity from clean and renewable sources including wind and solar so um we've got a 15 a 15 price to earnings ratio so right on target for us 116 um billion dollar market cap uh so it's a little larger than I think we were we were we were thinking of um but there's something good there I think it at least passes some initial criteria yeah what are you thinking um this is making me think that uh I just kind of keep going back to what works so well with Nvidia yeah and um I realize that there's the trading down due an due to an exogenous Factor stands to benefit from a technology wave and also there's this last aspect which is extremely scalable business model and uh I think that uh the the challenge I'm immediately imagining with some of these energy companies is that yes they're going to benefit from a technology wave uh but and we need to assess that specifically for each company right but then there's still the issue that they don't get this incredible scalability of the business model that a company like Nvidia or meta had okay got it and so now I'm a little bit stuck on how we're going to go about finding something like that in the energy industry and my instinct is let's ask okay cool I love that I think so I just want to point something out which is like we came into this with an idea of a trade we made a while ago and through the course of like using Gemini and talking about it um we and more specifically you have been refining the like key characteristics of this of the trade we we already made and the trade that we want to to make in the future and I think that's such a valuable process it's something that I think you're incredibly good at is really being able to articulate cleanly like the concept or the idea or the principle behind what you're doing and I think having that three-part trade so far maybe it'll maybe it'll get larger as we go is such a it's such a powerful thing I just want to like underscore that yeah yeah um it's trying to break down into what are the key components uh what are the makings of a company that's going to end up doing ridiculously well and uh seeing how much they can replicate in the public market yeah hopefully we we can find that it's a it's a high bar and like I guess it's it's possible that we come out of this not making an investment and some and sometimes that's okay um because we're really we are looking for that 10 bagger yeah holding your thousand in cash is uh not so bad uh you know still we'll still make 40 bucks on it yeah I'll put it in a high yield savings account exactly um Okay cool so uh okay so um help me um help me compose a a response to this uh we're looking for a company that has a particularly scalable business model uh so that as the technology wave uh of for example solar uh comes to to bear in the coming years uh this company stands to not only benefit but also scale massively um which companies in the energy industry that are public companies also have a highly skillable business model okay um so one thing that we should think about is like defining for Gemini like what you mean by highly scalable I think it's actually good to keep it think good keep it keep it out TR okay cool um okay let's we can always refine we can always refine that's that is true I think that's a good principle a start start simple in general and then refine depending on what we get yeah um I I find that using llms uh is a a task of continuously refining your question uh so it's seeing what's related to the way you ask the question and then uh you find some nuance and you kind of keep going and keep going and keep going totally um all right we're we're waiting we're we got the dots yeah I mean honestly I think that the biggest opportunity I've learned about today is hold music for LOL immediate investment okay um here are some public companies in the energy sector with highly scalable business models that stand to benefit from the solar technology wave First Solar is a leading manufacturer of thin filmer solar panels the company has a highly scalable manufacturing product and a global footprint as the demand for solar energy continues to grow for solar is well positioned nphase energy N nphase is a leading manufacturer of micro inverters which are used to convert direct current electricity from solar panels to alternating currents that's interesting that can be used in homes and businesses end phases microinverters are modular and easy to install which makes the company's business model highly scalable solar Edge Technologies solar Edge is a leading manufacturer power optimizers and inverters for solar energy systems like n phase Solar edg's Products are modular and easy to install which makes the company's business model highly scalable it's really keying on modular and easy to install which is interesting um Sunrun is a leading provider of residential solar energy systems the company has a strong track record of growth and is well positioned to benefit from an increasing demand for residential solar energy Sun Run's business model is scalable because it can be easily replicated in New Markets what do you think well um obviously manufacturers came out a lot here and uh I think you're right that we didn't Define highly scalable terribly well yeah uh but um if I think about again I keep going back to meta and Nvidia both of those have highly scalable business models in a different way which is uh you know as Nvidia is fabulous uh as a manufacturer so it doesn't literally have to go in manufacture the the items that just buys capacity from tsmc and uh you know similarly meta you know when when advertising spend is is growing and uh they're capturing more share of people's attention meta doesn't have to spend all that much more on on any important you know infrastructure thing or other thing that they they would need to spend money on um I think that these companies might have kind of the opposite characteristics and so uh I guess yeah maybe we should actually default since since we keep using it as inspiration maybe we should actually go flip back to asking Gemini and and chatu BT what companies uh have characteristics similar to meta and Nvidia do that and and Define some of those characteristics use that as I like that and I think we can I think we can list out our list out more details of our trade and Le list out details of meta and Nvidia and see what that does how that how does that sound okay so you uh you lead the way and I will I will be the Scribe uh we're trying to find a company whose business model is scalable in that uh it can grow a hundred times in a year and not uh uh run into any key bottleneck um that characteristic is extremely rare uh and more common in the digital world but we're looking for things that might have some similarity in non interet or software um in particular since energy seems like a great opportunity we're wondering about if there are any public companies in the energy sector that have this sort of extremely scalable business model uh um and at the same time uh sit in a a bottleneck in their value chain that um also in enables uh that that company to uh capture the Lion Share of uh New Market opportunity as it comes up you know in a in a Tailwind Market something like that W you can you can uh give give the llm basically War aibum text for slightly better total stream of Consciousness and it can just do something with it one of my favorite features it's so good right um Okay cool so um we're trying to find a company whose business model is scalable and that it can grow 100x in a year and not run into any bottleneck that characteristic is extremely rare and more common in the digital world but we're looking for things in non- Internet and software in particular because energy seems like a big opportunity we're wondering if any companies in the sector have this characteristic and sit in a bottleneck of the value chain that allows the company to capture the Lion Share of new opportunity as it comes up in a Tailwind Market I'm going to take this exact thing and put Pace into chbt and see if yeah let's do it we'll go ahead also we even invented the word tailwind and uh I know exactly what it means it's great it's a great word so does chat gbt Gemini I believe in uh finding a company in the energy so chat GPT says finding a company in the energy sector with a business model that can scale 100x in a year without running into any key bottlenecks is indeed a challenging task in the energy sector of physical constraints blah blah blah um energy storage solutions companies that provide Innovative energy storage solutions such as Battery Technology or Energy Management Systems might have potential to scale rapidly Energy Efficiency Services companies that offer services to improve Energy Efficiency such as smart grid technology could have scaling potential renewable energy trading platforms platforms that facilitate the trading of renewable energy or carbon credits could potentially scale electric vehicle charging networks companies that develop and deploy EV charging infrastructure might have potential to scale quickly cool uh let's try to get it to list some companies give us some companies please um and then okay so let's see let's compare to what Gemini said Gemini said it is extremely rare for a non- internet and software company to have a business model that is scalable enough to grow 100x in a year without running into any bottlenecks however there are a few energy companies that have the potential First Solar is a leading manufacturer of thin film it's really keying on First Solar um the company has a highly scalable manufacturing process in a global footprint they're more efficient and affordable it's well position to scale it also sits in a key bottleneck in the Solar value chain the company is one of the few manufacturer of thin film solar panels and has significant cost advantages while it is unlikely that for solar solar will grow 100x in a year the company does have the potential to scale its business significantly other um other companies are mhase which had already suggested solar Edge which it suggested and Sunrun which is already suggested so we're not getting anything like totally groundbreaking here yeah and um you know I just quickly looked up enas i' never heard about them before and uh it would have been an awesome trade for us in 2019 if we' found them at the same time we found media and and you know I think I'm not sure if today they're they're so unknown okay so we're running into the the challenge that um you know I think a lot of these are are probably pretty packed trades can we say like can you give us some more stocks that are less well known yeah and and also I think at this phase we might also switch out of the energy industry the only reason we got down deep into energy was because it gous shock exactly um maybe maybe one thing we could do is um is just start over with like the the the three or four part trade that we have so it's like exogenous shock um sits in a key bottleneck of the value chain and and is able to scale there's one other component that I'm forgetting um I think that those are the three major ones yeah um cool so I'm gonna just uh create a new chat prompt um I don't want to I want to save this cancel okay okay okay so we're looking to find public stocks that have the following characteristics are suffering some kind of exogenous shock to their business um sit in a highly valuable bottleneck in the value chain um and um uh have a highly scale scalable business model such that they could scale 100x in a year um easily relatively easily um anything else that you think we should put into this prompt all let's try that list them out um okay same thing we're going to do same thing for chat gbt and remember chat gbt has access to real-time information so I'm curious how it does relative to Gemini finding stocks that meet all three of these criteria can be challenging this is gbt's response um I can suggest a few Industries and companies that might fit these characteristics cyber security companies like crowd strike and pet alal networks that in a valuable bottleneck they have scalable businesses um they could face exogenous shocks from regulatory changes major security breaches or shifts in technology cloud service providers like Amazon ads and Microsoft I think we can get rid of that semiconductor companies like Nvidia I think we can get rid of that um biotechnology and vaccine manufacturers so companies like madna and bio entech have shown that they can scale rapidly and respons the global Health crises they sh they sit in a valuable position in in the value chain um blah blah blah uh fintech companies like Square PayPal and Aden so I don't know I think these are okay but um can you I'm just gonna ask can you dig deeper and find less well-known companies um and let's let's compare to Gemini so Gemini says shock proof stocks I don't know why it's it's uh so basically it's finding it's finding stocks that are in each of the categories instead of all three categories um so it's saying shockproof stocks um and and anad Darko petroleum Carnival Corporation Boeing I don't know if Boeing is shockproof um all it takes is one door to fly off of 737 match or whatever yeah um bottleneck shot stocks Google Microsoft Visa scalable stocks Amazon Netflix so it's giving us like really kind of generic suggestions um Can You Dig Deeper I'm just going to say can you dig deeper and find some less well-known stocks um so going back to chbt it says here are some less well-known public companies that might fit your criteria digital turbine this company provides mobile advertising and app Distribution Solutions the trade desk that's an IA company that's kind of funny um uh anari medical this medical device company special in treating Venus diseases um Ser specialized in creating unique moving experiences for the automative World upstart Holdings jrog matterport anything uh anything pop out to you um well uh seeing matterport show up uh just reminded me that um I believe matterport was an instance of a company that got public by being a a spacky company M and um I think the spec companies all got pretty wrecked yeah and uh are now uh um are now particularly small cap like some of them are probably barely staying on the exchange yeah and so um there might be an interesting opportunity lying in the the list of stacked companies that'd be fun to analyze that would be fun I'm also just noticing that um it doesn't feel like we're getting terribly useful uh responses here yeah um it's going to require quite a bit more digging yeah and uh I think it's almost like the same work that you'd expect an analyst in our little hedge fund to to be doing every day we're gonna get a ton of stock pitches that are no good so we've got to move our attention to a better opportunity next analyst please so uh I'm thinking about the the spack companies what do you think I think we can do that I think we can also so we've got about 15 minutes left um and I want to make sure that there's time to kind of like reflect on what we what we have experienced um and what we've learned from this um so I don't know uh are are you cool to to move on to that or do you do you want to push forward one more time I just feel like we got to make a bet okay yeah so we've got 15 minutes the clock is picking let's find our bet so we're gonna we're gonna bet how about this within 15 minutes we're going to bet on one company that has stacked gotten public and gotten wrecked in the public market and we're going to pick which one we should bet let's do it and do not do this at home absolutely do not do this at all okay so how do we how do we get all the the spack companies let's just ask for a list okay can you give me a list of all of the companies that spaed in 2019 was it it's 2020 to 21 2020 to 2021 um chat jbt says no okay well that certainly ended that conversation J I'm uh I'm trying to find some there's there's a website called SP track so I'm uh I'm just looking at the the spe track list let's see if can download it so I'm getting I'm getting a list so DraftKings nicoa Corporation I think that's gone now Virgin Galactic Open Door quantumscape charge Point Clover Health Sofi Lucid 23 and me we work grab Holdings bark AST space honestly like I think Open Door is an interesting one all right like they were doing really horribly for a while and I wonder where their stock is now um let's go back to our trading view um what's their ticker uh what's open doors ticker Open classic okay so looking at Open Door we've uh year-to date is down 33% over the year it's up for the last year it's up 51% five years it's down 73% market cap is 1.98 billion so it's very small um Open Door engages in the operation of buying selling and trading of Residential Properties online the company was founded by Eric Woo and it's headquartered in Tempe um it has a thousand employees what should we what should we look at uh well um starting to think about the most efficient way to cut through this because we only have 10 minutes right um so open door doesn't feel like uh one that I I would spend much time on um go back to the list okay so um I'm wondering about uh I think we should should take a bet on a company that has you know that got public but had some like pretty speculative technology and uh there's like a a bet that we could take on their speculative technology working and if it works the massive upside scalability of their model etc etc uh so let's see if we can find an instance of that here um you know what what stands out to me is um actually very few of these have that characteristic but it's also a very incomplete list um I'm interested in uh uh 23 and me I don't know if it's still public but I think it I I don't think so or there's some like really bad scandal with it right well there's a bad scandal with it so that's that's the exogenous shock uh and um I'm pretty sure that it's it's like it's still public it's trading at 58 cents there we go it's a $270 million company and um you know it's it's a bit of a of a Rat's Nest of regulatory issues and whatnot but uh you know potentially potentially interesting it's um at the the same time it's uh it's also got a uh um the characteristic of being sued a lot which which we we've gotten burned by in the past but I do think like genetic data is going to be way more valuable now with Transformers because you can throw it like you can process it all and make interesting predictions from it potentially yeah in a way that you couldn't do prior to Transformers and I really wonder if that data will be really valuable yeah and uh clearly it's h it's gotten battered to such a level that um you know there's there's definitely margin of safety if we end up being right right if it doesn't go bankrupt if it doesn't go bankrupt uh which I think is a key question is is is it gonna go bankrupt um so uh let's see well we've we've got we've got under 10 minutes we've got to figure out if they're gonna go bankrupt yeah uh and um and so let's uh let's try to figure that one out with uh with Chachi BT yeah let's do that um what what uh what would you what would you ask Chachi BT is 23 and me going to go bankrupt is 23 23 me going to go bankrupt okay um cool and it says as my last update in April of 2023 there are no public indicators no that's that's you're wrong um I'm GNA just add a new uh chat is 23 and me going to go Bank go bankrupt make make your best guess 23 me is facing significant financial and legal challenges but is not clear the company will go bankrupt the company's stock has fallen dramatically and is in danger being delisted additionally it's dealing with multiple class Asing lawsuits related to a data breach last year the company's financial situation is precarious with it with a substantial portion of its Venture funding already spent however it's attempting to transform into Healthcare company as some drug candidates in early stage human trials the future of the company will likely depend on its ability to navigate its legal challenges and pivot pivot its business model feels like a tough bet maybe we go back to our list all right going back to our list uh we've got Sofi we've got um I don't know what Quantum Escape is got charge Point yeah that's a that's a network uh oh man this is a tough list tough list um what's Virgin Galactic up to I don't know let's ask what's uh what is Virgin Galactic up to browse it recently completed its 11th successful space flight known as Galactic 06 on January 26 2024 this was notable as it marked the first time all four seats aboard were occupied by private astronauts it reached blah blah blah however Virgin Galactic plans to Halt its Unity suborbital flights by mid 2020 24 as it concentrates resources on the next generation of vehicle the Delta class ship the company will reduce the frequency of these flights and it's part of the transition to achieve positive Cash Flow by 2026 let's see what it stock price is doing Virgin Galactic um so uh let's see it's down a lot um but not really an exogenous shock it's just it's just it's just performing poorly yeah uh man this is this is tough I I I'm starting to feel like that 4% interest rate is looking good on the bank account um I mean I think we got to hold ourselves we got to hold ourselves to yeah we we we've gota we've gota make it we've got to make a bet we gotta make a bet I'm uh you know got that thousand um burning a hole in my Robin Hood account um I think we could also brought in to um if there are other stocks that you think are are good or or have been have been thinking about that don't necessarily fit this thesis no I want I want to do a dpack so or you know a battered stock so I think I think let's go with 23 and me great and I think we should flip from uh analyzing the stock to becoming a promoter of it so uh let's let's get Gemini to uh it listen to its to read it it transcripts and uh construct a a case for why it's going to uh be a 10 bagger for us I love it okay cool so so we're gonna get the um the webcast um all right let's try webcast sorry uh this is gonna be the same yeah the transcript is gonna be the same problem as the last time yeah um this that's the Q&A yeah I think that's that's fine the Q&A is okay okay um so we've got 2024 earnings call the Q&A and then uh we've got let's keep going there's probably some investor presentation that's worth it as well as yeah all of these other things okay cool uh all right so we've got a bunch of these oh wow interesting look at look at this this is like sort of a AI enabled Q&A um or maybe it's not AI enabled but it's very organized and it's not like a oh it's not a transcri I thought this the earnings call got it we've got the investor presentations but I think um yeah I don't I don't see that um available all right they might not publish me earnings call transcripts let's see if we can find it Seeking Alpha all right we've got it from Seeking Alpha thank you um so this is Q3 2024 um let's see oh 23 and me stock Falls 20% amid earnings report breakup speculations should probably read this before we we YOLO into it I can't read it because I don't have access to it yeah uh 2023 23 the internet's so broken now break speculation gole we've got three minutes all right I mean it feels like it feels like if the clock is ticking we've we've got to probably make a buy 23 and me consider splitting up company to revive stock price a day after the company reported dismo third quart third quarter uh 202 results and discussed splitting itself in into to help choose it stock 23 me reported 44.7 million for the quarter down 66.9 and it received a deficiency letter giving the company 180 days to bring it share price back and wsky said the company is considered splinting up its consumer in third peic business so feels like a bue to me this is the moment of sh so I mean I think at the end of the day we we've we've already decided that this is going to be our buy and now we're now we're just talking about constructing the investment these I think I think you're right um I I will say like I feel uncertain about this buy extremely uncertain about all the characteristics of all of our worst but I'm doing it for we're doing it for the content we're doing it for the fans we're doing it for the audience yeah exactly this is what they call this is what they call a YOLO trade um okay we're buying $1,000 worth of 23 and me all right this is our this is our stock that the AI has has recommended the funny thing is that the AI didn't recommend it at all this is the stock that we came to at the end of a long journey rejecting all stocks and I think I think so first of all I'm gonna press buy and I'll let you I'll let you see it as I do it we're going to swipe up to submit order received so there we go there we go we made an investment we made an investment I feel horrible about it y me too I don't think this one's going into our Hall of Fame I don't think so either but I think I think there's there is um a lesson here which is um when you when you have a really good strategy and I think we have a good strategy there's it doesn't there's not always an opportunity that fits that strategy and it's really easy easy psychologically to peer pressure yourself as we just did to make a horrible decision because you want to do something and I think that that's like really core to to the psychology of a good investor is you have to be really good at doing nothing um which most people are not um and uh like I actually think that if we if we sit on this thesis which we will because it's it's already always in our heads there will be over the next like couple years like one or two opportun that pop up that will obviously fit it and we will be able to take advantage of it yeah uh that really stands out to me uh when um when we saw that happen to Nvidia uh or you know meta stock price being depressed those were the fat pitches and you just have to be ready and and wait for them uh I I should know I I regularly reference Sports analogies in my life as someone who is not into sports at all and um and so I actually have no idea what a fat pitch is but I just know that that it's that I need to look out for and and this was not that this like exactly the opposite um my Reflections just since we were you know really trying to feature using LS today was um the massive context window is a huge benefit for Gemini yeah uh but the challenge with dealing with massive context Windows is that you need to find enough content to load into that context window to make it right and we just ran into time and time again over and over that the internet today is kind of broken and you can't really easily access lots of content largely I think because no one wants to have that content index by llms uh and so that's this kind of circular thing that I think is going to happen for a lot of the biggest benefits of llms in the near future is that you won't be able to access the content that most uh uh takes advantage of their context window I think that makes a lot of sense My Hope Is that big book publishers and movie studios and whatever will do deals with llms like the same way that Reddit has just just did a deal with Google to provide content um so that you can like you know instead of maybe me buying a book for me to read I could buy a book or like a set of books for my llm to have access to to ask questions about and I think that would be uh a world where you can you can have people who create content like benefit from using it inside of LMS and I think that'll be better for Content creators because people like it makes the books more useful it makes you like if you don't if you don't have to read a 300 page book like your LM can read it you might get value out of it that's locked in there that would be relevant to you but that you wouldn't know because you can't spend the time on it I was just about to make the same observation if if you think about how many products that we ran into just during this hour of messing around with llms and trying to make a stock pick every single one of those roadblocks that we hit we would have been willing to pay a Content site a a bit of money like buying an app in the app store to get that content into the m and make our next bit of progress totally and I think there's so many opportunities like that for Content companies on the internet 100% I'm like just envisioning like a one-click thing that it's like take all this stuff and just like like push it into like some some vast file store that like is hooked up to gemini or chat BT or whatever and you can just scroll around the internet like clicking and clicking and clicking and each time it's like or $3 or whatever and then you can create this like large personal database that you can use to um make decisions or find stocks or whatever yeah and I I think that uh that information being updated is another really key input so there were so many times that we ran into a roadblock that looked something like well we don't have the latest data or we don't have the the list of stocks and all the data required against them so um anything that has a real-time data source is going to be really advantaged that's something that stood out to me totally totally the other thing that's stood out to me was um aside from the uh information availability there's this uh issue where the LMS are not really willing to just make a recommendation and in fact it feels like they're getting worse and worse over time yeah uh there's the sense of they're just getting a little bit frightened of of saying anything too direct and um and so I think that the Golden Age of llm use might be in the near future when uh you know things haven't gotten lock down too far just yet and you're able to get enough context into that context window I think that makes a lot of sense there's also just this thing where there's always opportunities like if large scaled llms get to kind of milk toast it means that there's always opportunities for people to build ones that are willing to return riskier results um like that's the limits of the of the model's usefulness and intelligence is not actually technology driven anymore um like that is one limit but like mostly it's like what do we allow the technology to do and I think there's a lot of opportunity for people who are willing and can take the risk to allow it to do more stuff to um to make llms that people want to use yeah yeah we're gonna have to make another stock bet again in six months to see how it does Let's Do It um well thank you so much for coming on this was amazing I loved every second of it and um I'm I feel terrible about like that was the worst the worst way I've ever spent $1,000 but it was it was worth it to get to sit down with you I I I think it's worth it too and you know as someone who didn't uh spend the Thousand doll on the bet I still feel pot committed to it be watching 23 in me every day now amazing awesome um well thank you for this of course see you next time after after that episode like I just could not let the 23 Me by stand and and I was like thinking about it and I think you were thinking about it and um I was watching the market and I think I found something that sort of actually fits the thesis and what I want to do with you is one we actually made some money already on the 23 and Mi bu so I want to I want to exit that position I want to I want to take the win and then I want to go into the into another stock that I think like uh fits the thesis and I want to help I want you to help me explain why it might fit this thesis uh for everyone who's who's listening I can't believe we made money in that trade but let's do it uh so this kids is how you make money on a day trade um uh we are uh we're we're down minus 4% on the 23 and me since we bought it uh last week and we're up 9% total so I'm just gonna I'm gonna sell all of it uh I'm gonna do a I'm gonna do a sell all trade um we're gonna make some money uh and done is it done it's done we made money okay wow consider considering how little thought went into that that's uh that's actually pretty impressive that's pretty good you know a a a stop clock is right twice a day or whatever you know um I feel I feel pretty excited about it so the thing that we want to do now is I actually think that uh Google sort of fits this thesis to some extent um it had it had uh like a poor it had poor earnings uh like a month ago but the real the real interesting thing that occurred to me is they've had this sort of exogenous shock where over the last week or so there have been all these problems with Gemini and I think that that is driving down the stock price like it's lost 70 billion dollar in value over the last week and I just feel like it might fit that thesis what do you think yeah I I think that Google is uh I mean it's it's up 50% year-over-year so it's it's hard to say that it's uh you know do doing poorly but but um but in the in the last few weeks in particular um there was the combination of a negative perception moment um which they've been affected by a couple of times before especially on the sort of comparative chat GPT uh and um and it was combined with their earnings disappointed because um their revenue for ad uh impression wasn't as as good I think as well as um their capex was high unusually high and uh I think investors are are right to perceive that well s something is changing in the Google model they're going to have to spend more on capex for longer uh in order to serve uh uh search with Gemini uh and um with llm type features to Ser cost more to deliver uh and at the same time it's harder to load ad units into it so the the um fears about llms were confirmed by the recent earnings even though obviously they're not terribly affected by LMS just yet uh and um and at the same time people's fears about Google being incredibly corporate uh woke unable to get out of its own way were confirmed by all the uh uh the uh Twitter chat chat gbt Dolly comparisons to uh asking questions of Gemini um so I I do think that that this past couple of weeks would qualify for the exogenous shock component of the thesis uh there's still the other parts though too which are well does the company have a Tailwind that it benefits from and does it have an extremely scalable business model so that when that Tailwind shows up the company's able to capture a lot of the flow of demand and um I think it's worth repointing out that Google has one of the most scalable business models ever conceived uh so uh so that helps but if we specifically analyze Google uh from the perspective of uh well in the future they're going to need to deliver search results uh in a way that requires a lot more compute uh they are still advantaged in that business model because inside of Google they've built their own basically their own mini media uh they they actually design their own semiconductors uh and then they they have their own cloud uh to deploy those uh and uh deliver on aib based search as well as cloud services uh and and so their extremely scalable business model in spite of having a bit more cost in the near- term is actually a huge beneficiary of this Tailwind uh in llm generated search uh and uh llm generated experiences so uh I do think that it it qualifies uh so long as you're okay with the lower margin of safety uh with with you know it's not it's not like the stock is way way down you're right you're that that is true I I I I am because one of the things that I feel about Google is you know people are really down on Google they're like search is going away blah blah blah all that kind of stuff I think it's reflected in the stock price but I think people forget that everyone felt that way about Microsoft like or 10 years ago and before that everyone felt that way about Apple like 15 years ago and both of those companies managed to get through those periods and come back really really strongly in the subsequent technology waves and I just feel like Google may be going through a rocky period it may go through a rocky period for a while longer but I do think it's well positioned to grow a lot in the AI wave I mean the fact that they uh they got you to start experimenting with Gemini alone I think is the best indicator uh and so if they're able to keep your attention and you find yourself using Gemini more and more I think that's a good reason to to hold on to it I agree okay so let's let's roll our 23 and me winnings into Google and hope that Sundar delivers um so I'm going to do dollars and I'm going to put in 100087 so we're buying we're buying uh 1,087 worth of Google and we're heading by we did it we're in Google go Gemini go Gemini uh go Sundar uh make us some money thank you Jesse for for coming on and completing this last segment of the episode I'm excited for it and I'm excited for our next 10 bagger all right so excited to to see what happens here thanks a lot of course oh my gosh folks you absolutely positively have to smash that like button and subscribe to how do you use chat GPT why because this show is the epitome of Awesomeness it's like finding a treasure chest in your backyard but instead of gold it's filled with pure unadulterated knowledge bombs about chat GPT every episode is a roller coaster of emotions insights and laughter that will leave you on the edge of your seat craving for more it's not just a show it's a journey into the future with Dan shipper as the captain of the spaceship so do yourself a favor hit like smash subscribe and strap in for the ride of your life and now without any further Ado let me just say Dan I'm absolutely hopelessly in love with you