Early Retirement Extreme Forums

Getting into particular stocks is also a bit off-topic to the thread, imo. I propose we keep this focused on the broader topic rather than specific stocks, maybe those can be discussed in the investments section of the forum?

We are all dealing with the entry of artificial intelligence into our daily lives due to its development and commercial deployment. To not track the on-going development and commercial deployment of it in a conversation focused on its future seems, I don't know, stupid? Yes, I am coming at AI strictly from the perspective of how can The House of Henry benefit from it. It's like the group of cavemen sitting around the first campfire staring at the flames. Jerkhoffth is thinking "Me going to open Jerkhoff's Barbecue Stand". Kockthrust is thinking "Me thinking this will make Krackhoor horny." Krackhoor is thinking "Me thinking Kockthrust pay to fuck me by fire." Zykopath is thinking "Me now burn down Jerkhoff's Barbecue Stand." MaulKopp is thinking "Me thinking we need water to stop Zykopath from burning down Jerkhoff's Barbecue Stand," Brainzfark is thinking "Me thinking Jerkhoff pay me monthly in case Zykopath burn down Jerkhoff's Barbecue Stand." Bigheadth thinking "Me thinking this change civilization." You get my point.

Henry wrote: ↑

Sat May 31, 2025 6:08 am

... Jerkhoffth is thinking "Me going to open Jerkhoff's Barbecue Stand". Kockthrust is thinking "Me thinking this will make Krackhoor horny." Krackhoor is thinking "Me thinking Kockthrust pay to fuck me by fire." Zykopath is thinking "Me now burn down Jerkhoff's Barbecue Stand." MaulKopp is thinking "Me thinking we need water to stop Zykopath from burning down Jerkhoff's Barbecue Stand," Brainzfark is thinking "Me thinking Jerkhoff pay me monthly in case Zykopath burn down Jerkhoff's Barbecue Stand." Bigheadth thinking "Me thinking this change civilization." You get my point.

I don't, but it's clearly very valuable to the discussion.

Since my last post was illegible, however, I will state that the reasons I think it is possible Tesla is in the Blackberry position rather than Windows position is that

1. cars haven't had a network effect in the past and so I contend moves to make artificial scarcity will be transparent and resisted
1A. The very person at the head of the company has pissed multiple nations and states off. He will not be granted artificial monopolies.
2. I would bet on open standards, at least them existing and being viable. I don't think Waymo is Jeeves, but pronto-Android.

Yes, Americans can give 57% marketshare to Iphones, but that is our current wealth level and on the back of multiple apps and friction to switching. I don't think it is just sleekness of appearance. And globally speaking, Android is the marketshare leader. "First in, takes all," makes no sense, at least globally.

candide wrote: ↑

Sat May 31, 2025 7:56 am

I don't, but it's clearly very valuable to the discussion.

Since my last post was illegible, however, I will state that the reasons I think it is possible Tesla is in the Blackberry position rather than Windows position is that

1. cars haven't had a network effect in the past and so I contend moves to make artificial scarcity will be transparent and resisted
1A. The very person at the head of the company has pissed multiple nations and states off. He will not be granted artificial monopolies.
2. I would bet on open standards, at least them existing and being viable. I don't think Waymo is Jeeves, but pronto-Android.

Yes, Americans can give 57% marketshare to Iphones, but that is our current wealth level and on the back of multiple apps and friction to switching. I don't think it is just sleekness of appearance. And globally speaking, Android is the marketshare leader. "First in, takes all," makes no sense, at least globally.

My point is that there is a multiplicity of views towards everything, specifically a tectonic, or potentially tectonic shift in technology. Warren Buffet's response to the arrival of AI was to highlight its potential nefarious use, which we are seeing ie increase in scamming. Elon Musk's attempt at totalitarian industrial control needs to be evaluated in terms of ethics - does he have the best interests of humanity in mind. I think he does. Some college kids are trying to use it to get out of writing papers. I just want my portfolio to increase. Some scientist is thinking can it help him solve cancer. We are all staring at the same thing from different perspectives.

I don't understand your point. You first are saying there won't be a network effect. OK. I think there will be. But then you say there will be open standards and Elon won't be given a monopoly. Why is there a chance of Elon gaining a Monopoly if there is no network effect. Why are you going into the Android/Iphone and US/World debate if your premise is there will not be transportation revolution in the first place?

Henry wrote:Bigheadth thinking "Me thinking this change civilization." You get my point.

Yup. I think questioning the motivation of others is increasingly besides the point. The development of AI is much like climate change in that there is no existing power structure adequate to address it. The CEO of Anthropic, Dario Amodei, when recently questioned about why he continues to contribute to the development of AI given the associated risks he was attempting to share with Anderson Cooper, replied "Because China." Which is exactly the reply I received from my multi-multi-millionaire friend when I questioned him about climate change.

One interesting thing about operating systems is that it has been proven that they must inherently suck for pretty much the same reason governments must inherently suck. They are engaged in the distribution of limited resources. Within the current framework of an AI agent , the AI (or AI access) is roughly analogous (and also somewhat literally) the processor. Assuming you aren't a genius of design, ultimately the only particular value to be derived from a personal AI agent will be based on your proprietary information. It's not difficult to imagine a world of the future in which personal AI agents chat with each other like nannies on a park bench in order to arrange play dates for their humans based on the only remaining purely individualistic human proprietary information which would simply be our subjective preferences. IOW, not unlikely horror-show endgame from the perspective of a member of this forum is that AI will transform all of us into nothing but consumers.

7Wannabe5 wrote: ↑

Fri May 30, 2025 4:27 pm

I have been using Plant ID apps for many years and they have become increasingly reliable

What stood out to me, is I'm using a general purpose agent with chatGPT. I didn't get or learn the plant app. I didn't even constrain the agent with a role or a goal.

Instead - I asked very low context questions. What are these spots? What do I do about it? My error check - what type of plant is this?

If the AI has broad intelligence, it becomes increasingly rewarding to lean into unstructured requests. Maybe it can extend "I want a thriving deck garden" to a proposal for approval, with no further interaction required.

It's a potentially huge disruption for existing specialized software. Instead of a plant app, a money app, a health app - chat GPT could replace it all. It doesn't even have to be better, just good enough to force the specialized products up market, as described in the innovator's dilemma.

As consumers, we see only one class of this software. There are specialized applications in business as well, often being sold at exorbitant prices, framed as software as a service. What if the general purpose agent can out compete those?

The specialized apps augment or substitute for professional judgement. When we can skip the entire development curve for individual domains, wild stuff happens.

Given the rapid pace of change - as an individual, developing specialized skills against the intermediary solutions offers rapidly diminishing returns. They might be obsolete before we're even up to speed.

7Wannabe5 wrote: ↑

Sat May 31, 2025 8:22 am

One interesting thing about operating systems is that it has been proven that they must inherently suck for pretty much the same reason governments must inherently suck. They are engaged in the distribution of limited resources.

Or, possibly in the instance of AI, the inverse is true - the operating systems could be engaged in the distribution of unlimited resources, but are currently limited by both governmental power and electrical power. Elon's engagement in DOGE was to inform the government that AI could potentially provide unlimited amounts of data but is restricted by the current US administrative state and the current US power grid. Energy is the least discussed but possibly the most significant issue in the Future Of AI discussion outside of what color Candide's open source made in China EV will be. Elon is tweeting "terrawat" as describing the energy needed to support his AI visions. I am not an energy expert, but I have heard secondhand that's like a full US boatload of power.

Scott 2 wrote:If the AI has broad intelligence, it becomes increasingly rewarding to lean into unstructured requests. Maybe it can extend "I want a thriving deck garden" to a proposal for approval, with no further interaction required.

Yes, my thoughts exactly, and I would extend to note that level of delegation is also arbitrary, so the submission of proposal for approval step could also be skipped. By analogy, when I was running my book business, at one point I completely delegated the shipping department to my daughter and one of her teenage friends and the warehouse department to my son and one of his friends. So, the most frequent sort of communications I would get from them were to do with shortages of flow of resources such as packing envelopes or storage boxes. If I had given either of them access to the accounts through which I ordered these supplies, they could have simply done it themselves with no need for communication.

Given the rapid pace of change - as an individual, developing specialized skills against the intermediary solutions offers rapidly diminishing returns. They might be obsolete before we're even up to speed.

Well, I'm about the last individual to be motivated by the prospect of developing specialized skills so that I can sell them on the full-time employment market. I'm more interested in performing experiments to gain insight into this expansion of the map of technology. I think this is worthwhile even if analogous to being at the level of "creating a herb spiral" by following step-by-step instructions in the realm of permaculture. IOW, even though I definitely err on the side of "learn by reading", I am personally at the juncture of "learn by doing" in the realm of AI. Actually, I landed at this juncture back in October as I was wrapping up my M.S. in IT/Data, but I needed to relax/warm-up/relate in my feminine-energy-mode for a bit before coming back to it. I should be utilizing my tiny-clit-energy* to whip myself back into shape fitness-wise, but building muscle on my body seems kind of boring compared to building an AI agent.

*female-based masculine-energy locus. for information pertaining to male-based feminine-energy locus, see episode 5, season 2, "Sex and the City" series sequel.

Henry wrote: ↑

Sat May 31, 2025 8:10 am

Why are you going into the Android/Iphone and US/World debate if your premise is there will not be transportation revolution in the first place?

That's not my premise. How are you parsing the sentence in an earlier post "The critique wasn't about the possibility of technological changes coming, it was your hyper-specific scenario involving one company you are being a stan for."

At this point, I am just going to turn it over to GPT to try to explain what I am saying (one reprompt to get my output):

You're misunderstanding my premise. I'm not saying there's no transportation revolution. I'm saying that even if there is one, Tesla isn't guaranteed to dominate it. The reasons are:

Cars don't naturally accrue network effects in the way software platforms do. Just because more people use Teslas doesn't automatically make them more valuable to others. A car isn't Facebook or Windows—its value doesn't scale socially in the same way.

Attempts to create artificial scarcity or platform lock-in will be resisted. Moves to wall off charging networks, data ecosystems, or self-driving software will face pushback from consumers, regulators, and competitors. That makes the "Tesla-as-Windows" scenario unlikely.

Elon Musk has alienated multiple governments and regulatory bodies. The idea that he will be granted monopoly privileges or regulatory advantages is implausible, given how many institutions he's antagonized. That matters when you're trying to create a dominant infrastructure or standard.

Open standards tend to win, especially on a global scale. Even if Tesla builds a lead in autonomous driving or infrastructure, other players will emerge, and standards will form—just like Android did in the smartphone world. Waymo, for instance, is less like Jeeves and more like the foundation for an open alternative.

The iPhone/Android comparison is there to illustrate that even in a domain with network effects and switching friction, the open platform (Android) wins globally. The U.S. market is anomalous in giving 57% share to iPhones, but that's driven by wealth, app lock-in, and cultural preferences—not just sleekness. Extrapolating U.S. consumer behavior to the world is a mistake.

So again, I'm not denying technological change or a transportation revolution. I'm critiquing the narrative that Tesla (and Musk) are destined to dominate it in the way Windows once dominated PCs. That argument relies on assumptions about network effects and monopolistic advantages that don't hold up under scrutiny.

(I then kept the conversation going, if you want some bonus content that is off the topic, but potentially interesting).

candide wrote: ↑

Sat May 31, 2025 1:53 pm

Elon Musk has alienated multiple governments and regulatory bodies.

Outside of California, who is protecting Waymo's State Constitutional rights to take perpetual right hand turns around San Francisco, who exactly are we talking about here? I mean I just saw the Head of the US Department of Transportation in his Giga Texas Factory. I believe the President of The United States invited Elon Musk to contribute to many facets of US domestic policy. I don't recall the British Parliament or the King of Sweden saying they dislike Elon Musk.

Furthermore, which US car manufacturer possesses the technology as well as the ability to manufacture at scale that is even close to TSLA at this point? Seriously. Name the US manufacturer that is a close second? I'll go further, who has the prototype to compete with TSLA?

For purposes of clarification, when I read your anti-Elon bias that is Chat GTP? But when I read your anti-Henry bias, that is you? I'd prefer Chat GTP if that is possible. AI Hominem > Ad Hominem.

Henry wrote: ↑

Sat May 31, 2025 8:10 am

Why are you going into the Android/Iphone and US/World debate if your premise is there will not be transportation revolution in the first place?

I think the point is that not all revolution make stock go to the moon. In tech, there were several revolutions which led to natural monopolies (Microsoft, Apple, Google, Facebook) due to network effects and software (in)compatibility. Those companies are worth trillions now each.
In cars, that may not be the case. Just look at the weight loss revolution - Ozempik was the hit, but now there's plenty of competitors and Novo Nordisk is already down 50% from the top. Turns out, they had no moat. Who knows how it will play out for cars.

zbigi wrote: ↑

Sat May 31, 2025 3:07 pm

I think the point is that not all revolution make stock go to the moon. In tech, there were several revolutions which led to natural monopolies (Microsoft, Apple, Google, Facebook) due to network effects and software (in)compatibility. Those companies are worth trillions now each.
In cars, that may not be the case. Just look at the weight loss revolution - Ozempik was the hit, but now there's plenty of competitors and Novo Nordisk is already down 50% from the top. Turns out, they had no moat. Who knows how it will play out for cars.

TSLA is worth 1 Trillion.

God damn, I repeat - name, please fucking name, the auto manufacturers that are going to compete at this point. Who is close to ramping up the production of thousands of autonomous vehicles. Please. Name them.

BYD is out producing Tesla in vehicles equipped for autonomous driving, and is also eating Teslas lunch in every market they compete in? And is eating Teslas lunch in battery research? And is totally vertically integrated, and out produces / out ships / out prices / out sells Tesla? So what edge does Tesla have? I guess more sycophants, but that’s a net drag, not boon.

Tesla can maybe maybe win the US market, and that’s it. But they’re losing market share in the US. Claiming something is good because it has a high market cap is hilarious, especially when they’re selling a hype train while sales / market sentiment is cratering.

Henry wrote: ↑

Sat May 31, 2025 3:25 pm

Who is close to ramping up the production of thousands of autonomous vehicles. Please. Name them.

I doubt Tesla is, year after year we have had lies about how capable their self driving technology is.

Henry wrote: ↑

Sat May 31, 2025 3:25 pm

God damn, I repeat - name, please fucking name, the auto manufacturers that are going to compete at this point. Who is close to ramping up the production of thousands of autonomous vehicles. Please. Name them.

What is the chance that the details about minor (or even significant) advantage in this or that will have affect on position of car companies in 10-20 years? So much can happen in between, I don't think we can use the present to predict the future that far with this much certainty (and it's required to justify very high Tesla valuation). The history teaches us that early movers as often as not squander their early advantage and end up in the dust.

On the other hand, if you mostly care about just TSLA reaching $1000 for a split second so you can dump your stock for a big payday, that is much more likely to happen IMO. The stock is basically rallying point for the techno-optimist speculators. Enough short-term good news could push the stock way up for a while.

theanimal wrote: ↑

Sat May 31, 2025 2:21 am

@Henry has made it abundantly clear throughout his posts that he is long NVDA and TSLA.

Joe Bhakdi is a heavy TSLA investor and AI Doomer. He believes that starting within the next 12 months, NVDA will come under intense scrutiny to curb its compute. For now, it's a US vs China arms race so its go go go. But shortly, a simultaneous concern will grow to limit compute as it surpasses human intelligence. The bots won't just do your dishes but will possess AI Chatgpt compute power. What is the greatest threat to TSLA's autonomous driving takeover? Robots driving cars. He thinks the inflection point will be in 2027 when government steps in and says AI compute power needs to stop in order that we don't lose control of it. When that happens, goodbye NVDA. TSLA will drive on with its accomplishments. So, I'm probably longer TSLA than NVDA. Shit is getting weird and is going to get weirder real fucking fast and it's going to frighten the shit out of people. Grandma will love her FSD but when her robot gets sick of her there's more tea in the teabag nonsense and goes 2001 Space Odyssey on her ass and tells her she needs to shut the fuck up or it's going to put her wig first into her sugar bowl, she might ask you to take it back.

Round up of image and video generation progress:

https://www.oneusefulthing.org/p/the-re ... -32-otters

The author's success with local generation is eye opening. I still think of AI as something called from someone else's data center. But his offline results are surprisingly good.

My attention tends to center on the race towards possibly super human AGI. It's sensational. Pocket models running entirely on our phone though? That's an entirely different form of disruption. I was incorrectly dismissing the potential. Confusing training compute costs with the much smaller transaction compute costs.

I have a friend that runs a llm localy, the answer take a few minute instead of seconds, but they are a thr level of insight that a younger peer could provide. By younger peer, he means a fresh out of college engineer.

Henry wrote: ↑

Sun Jun 01, 2025 2:28 pm

The bots won't just do your dishes but will possess AI Chatgpt compute power.

The humanoid bots industry seems still in its nascency. AFAIK they don't even have a reasonable benchmark for quality of their robots - something like e.g. get into this abandonded warehouse full of cluttered mess, sort and arange things in the warehouse without breaking anything. The challenge would have to be set up by an external organization (perhaps a consortium of manufacturers) to make sure robot is not pre-trained and other cheats are not employed. Right now, all I'm seeing are worthless demos like robots doing backflips, or, at best, demos of robots seemingly acting intelligently in a complex environment, but all done by manufacturer, which means the demo is most likely a fake.

I can't believe we'll go from barely anything to T-800 in a matter of couple of years. Engineering doesn't happen that quickly.

Scott 2 wrote: ↑

Sun Jun 01, 2025 6:02 pm

Pocket models running entirely on our phone though? That's an entirely different form of disruption.

Ropes you into their ecosystem while increasing hardware requirements, forcing the need to upgrade. Makes sense it'll be something they'll pursue.