The pyramid scheme of raising money each year to avoid self-bankruptcy of billions of dollars worth of inference and training costs a year now must expand to investors in the middle east. (As US investors are almost out of cash and averaging themselves up.)
This scheme will soon become unsustainable like the last time the tech industry corrected in 2022.
Hopefully the participants don't experience any down-rounds or devaluations along the way. (It happened to many over-valued startups like Klarna, and Stripe and OpenSea.)
While I do believe the ever increasing “more money for a trillion dollar ai company scheme” is not infinitely viable, in what world are us investors running out of money?
> Hopefully the participants don't experience any down-rounds or devaluations along the way. (It happened to many over-valued startups like Klarna, and Stripe and OpenSea.)
Klarna is a debt as a service company. They were made to exist in an easy money world.
Opensea is a fraud as a service company. They existed in the crypto (more specifically nft) craze.
Stripes down round was due to changing economic winds, nothing due to the softening of their business
I thought the agreement is Microsoft gets full access to all the being models, except when they develop AGI.
If that's the case, I could see their relationship turning sour. OpenAI wanting to declare AGI to break ties with Microsoft; and Microsoft doing everything it can to deny that OpenAI has created AGI.
It is already there. Microsoft has intellectual property rights any pre-AGI innovations OpenAI makes, and most of the investment they have paid was in the Azure credits iirc, so basically in company scrip. And MS helped Sam oust rebels in the company and return. I would honestly consider them a part of MS now.
The Middle East is desperate to get out of oil. Heck, UAE has a Ministry for AI. And Saudi Arabia already gave a hundred billion to Softbank to burn, I'm surprised they haven't Khashoggi'd him yet.
It's difficult to get your arms around the absurdity that is the OpenAI corporate structure. A non-profit parent entity with a wholly owned for-profit subsidiary (actually, the non-profit wholly owns and controls a third "manager" entity which in turn is the controlling but not majority shareholder of the for-profit subsidiary) where the profit of the for-profit entity is "capped" and oh by the way, that for-profit subsidiary has a 49% shareholder who happens to be Microsoft. However, Microsoft, despite having 49% ownership of the for-profit subsidiary, gets 75% of their profit..........until their initial investment is recouped. Employees get Profit Participation Units (PPUs) which are NOT equity but which entitle employees to a percentage of profits, again up to the "cap". Then of course, when OpenAI achieves AGI, whatever the hell that is, then everything from that point forward reverts back to the non-profit entity and Microsoft, other investors, and employees get hosed.
I wish they would be at least forced to change their name. It’s like if Exxon called themselves The Free Renewable Energy Company. It’s misleading to consumers. In tech, we all know it’s BS, but the average consumer might not. It seems like an argument could be made that “OpenAI” in violation of FTC truth in advertising rules.
It could be a lot more closed than it is, from a general public perspective. i.e. the company having this cool AI and we not even getting the opportunity to use it at all.
They couldn't be, realistically, any more closed without being an essentially infinite money sink of the kind even nation states would struggle to fund long-term. It's the public using it that's helping offset costs and also giving them the usage data that they need to iterate on the design.
Yeah but even then it would make it to consumers through their customers.
A bit like MS Copilot for Bing / Windows which is really just rebranded ChatGPT. For the O365 version they added a lot of stuff to interact with office but their online chatbot is just the same.
Also, OpenAI is now one of the most closed AI companies. No open source. No open weights. They don't even share how they're models are designed or the number of parameters. There's just nothing left from their original mission.
There are only so many places you can put your money--treasuries, bonds, real estate, equities, precious metals, crypto. With interest rates going down, fixed-income looks a lot less interesting, so... real estate? Equities?
As someone who remembers that time very well I disagree.
Smartphones followed a similar path to the internet i.e. completely new paradigm, that had immediate benefit but needed technology to improve to become widely adopted. But where the roadmaps to improve these technologies were clear and well defined.
AI is far more akin to nuclear fusion where each step along the way will require major scientific breakthroughs. And where it's not clear how to get this grand AGI end-game. Especially as OpenAI has said advancements depend on compute capacity that we simply don't have.
Even with their current capabilities, these AI systems will dramatically improve productivity. We could have a 20-year AI 'winter' with no new advancements, and they'd still be a big deal.
The thing is that it will take years to integrate them into existing domains and workflows. Honestly I think the most relevant comparison is the rise of desktop computers themselves. Suddenly paper-based processes had to become electronic and in some cases that took 40 years.
Except we've had AI systems for a while now and it hasn't meaningfully impacted productivity across the economy. Maybe in select pockets e.g. knowledge workers and even then it's highly debatable.
You compare this to the internet or smartphones and they have been transformative across every aspect of society.
And as someone who works for a bank which is heavily exploring LLMs there is no complexity in integrating them into existing workflows. The issue is that (a) risk of privacy/security being compromised through prompt exploits and (b) risk of reputational damage if the prompt is biased or hallucinates. Issues that may well be inherent to all transformer based models.
The other main difference that I can’t get past is that the Internet and cell phone were so obviously useful to a layman that you didn’t really have to explain the value proposition.
The ability to type an email for instance, and send it instantly to someone on the other side of the planet for free was something that was in such contrast to the tools of the time it can’t really be understood by someone who didn’t live through that period.
AI to me is really hyped up by some very highly regarded CEOs with strong track records in other domains and tech enthusiasts who seem hell bent on being able to look back and say they called the next Industrial Revolution. In short everyone thinks they’re the counter to Paul Krugman saying the internet would be as useful as the fax machine. Credulity levels are off the charts. It’s gotten to the point where skeptics are automatically assumed to be wrong.
But what’s missing is the obvious amazement that should come to an ordinary person and frankly I still don’t see most people naturally gravitating to these tools.
Perhaps that could be explained by how amazingly fast technology has advanced in recent years but then that in and of itself seems to call into question whether this technology that’s being called AI is truly revolutionary when compared to what’s already available.
> I still don’t see most people naturally gravitating to these tools.
I barely know anyone who doesn't use ChatGPT frequently, to help wording an email or such. I agree though that in instances like this it is not transformative to society and rather one more tool that we use. We will see, IMO the impact of the current AI technology on the world is rather "medium".
ChatGPT feels like a more advanced Grammarly. In our company, they keep creating chatbots tailored to specific domains, but with poor data quality, the ROI remains low. Right now, it's mostly hype, and a true AI revolution seems years off. Even internally, executives are questioning how to measure ROI when they see the costs. I suspect the current hype cycle could lead to the downfall of some companies that focus too heavily on developing AI features for their stakeholders.
I think you may have a bit of hindsight bias. The internet was not immediately recognized as useful. One of my first jobs was at an e-commerce startup that folded in the dot com bust because (1) investors did not believe that our target customers would favor online shopping over ordering from physical catalogs, or (2) foresee how online sales might impact business operations or the viability / scalability of business models - such as drop shipping.
I think that you are not appreciating how much of a quantum leap this is in technology. This is like a calculator for verbal reasoning. Humanity has little idea of what is even possible with something like that.
Also, I am old enough to remember - and have experienced in my professional career - a time when the mainstream opinion was that the internet was a tiny niche and that things like online shopping or dating would never gain widespread adoption.
I LOVE it when companies replace their standard customer service with AI. It makes it much easier to get ahold of a real person, faster, by just confusing the AI system until it falls back onto a human.
Since you’re saying “for a while” I presume you mean ML. ML systems have had an enormous impact on productivity. There’s a huge volume of decisions that get made instantly and with far greater precision using models That were done manually before.
I've worked with ML for a while and I wouldnt say that this is the case at all.
ML also didnt replace manual decision making. A lot of previously automated decision making was done with what were basically encoded rules of thumb which didnt overfit much worse than "more advanced" ML models did.
Your personal failure to use ML successfully has no bearing on its wider deployment and success. Overfitting strongly implies that you or whoever was doing it just didn’t know what they were doing.
Dozens. I’m a consultant and it’s what I do. The modeling is very easy. The operational changes are the hard part. Basically any random task of. A sort of maintenance, production, replacement, procurement or scheduling task can save 10-30% with a simple model. There’s a long way to go but this stuff is everywhere.
The world hasn’t even fully realized the productivity value of spreadsheets and email.
The numbers aren't public, but it was widely reported earlier this year that their revenue is in the billions. I would assume this is mostly API use by businesses, not individual ChatGPT subscribers.
You won’t necessarily notice it as much in end user application. But I already see good adoption for knowledge workers and the next step is in process automation through agentic workflows. Most likely the result will be in the form of efficiency gains, improved profits and worse customer service.
I think it's more like nuclear fission, frankly. I don't think the major breakthroughs will be as difficult to achieve, but we'll have a trail of nasty side effects and bad actors that will be left in its wake.
In 2020 I never thought an AI would code for me. And now it turns out, it's pretty good. But AI Slop is pretty much a toxic mess.
> AI is far more akin to nuclear fusion where each step along the way will require major scientific breakthroughs. And where it's not clear how to get this grand AGI end-game.
We have AI now. We've had it for ages. We just stop using that term for things once we've gotten them working.
We also already have artificial entities that are smarter and more capable than a lone human, and have for a long time. Bureaucracy and writing are incredibly powerful technologies, especially when combined.
>AI reminds me of the internet in 1999." -- Peter Thiel[1]
It just feels like we're still in the pre-Navigator days though, circa '94. All of this complex orchestration work needs to be baked into a single model. I don't want to learn how to implement a multimodal agent system. I want to tell something what to do and have it do it for me perfectly with no more interaction than a simple prompt. Fortunately this is just a software engineering problem now, not a CS one.
I agree completely. The companies I'm currently most excited about are all working on making some sort of E2E platform, weaving together all of the component pieces into a cohesive whole.
Open Ai must be making some mad earnings that are not disclosed. Huge contracts I suppose. Meta was worth $100 billion pre-IPO but they were obviously very profitable and had lots of advertisers. It reminds me of Tesla. During 2020-2021 the stock went up 20x and it seemed irrational and the biggest bubble ever, but fast-forward to 2024 and Tesla cars everywhere especially in the Bay Area. The market correctly anticipated Tesla's popularity.
Tesla sells like 1% of the cars and has like 50% of the market cap, and it’s not because their margins are so high.
All the market correctly predicted is that the market would continue to incorrectly predict their value.
As a mental exercise, imagine there’s a bubble that went on for 5 years. They sometimes go that long or longer. Do you not think 3 years in people would think they were seeing the market just having been correct the prior few? It would look exactly the way you view this.
This has been a phenomenon of the endless liquidity train of the past 3-4 years. People doing very stupid things with money knowing more is always coming. There are certain stocks like palantir that are very obviously in bubbles where the market celebrates its inclusion in a stock index and sends the shares up 15%. The CEO just self produced a video recognizing retail traders and using the term “tendies.” The mistake this cycle though has been to assume that won’t work
> All the market correctly predicted is that the market would continue to incorrectly predict their value.
I don’t understand why so many don’t get this, especially after crypto currencies became a thing. The token can be DOGE or TSLA, the mechanism is the same, only marketing and PR content change, reality doesn’t matter. In fact Musk already said Tesla is not a car company anymore. Crypto opened the eyes of a lot of smart CEOs, and not in a good way, and investors look away as long as the dance continue.
I don't touch crypto, but one thing I took away from its rise is thinking through what money actually is and what equity valuations actually represent.
I’m not saying the should be valued higher. But it is at least notable that a Toyota has been the best selling car every year for the last 20 or so, until last year when it was the Model Y.
Meanwhile Toyota has been caught unsure of if they should build gas, electric, or hydrogen cars. Fortunately, as a fan of Toyota’s cars, it seems like they’ve figured out a good strategy forward. But for every Toyota there is a BlackBerry, IBM, Reebok, or Blockbuster who doesn’t figure out how to adapt to a changing market in time.
You are using a pretty deceptive number when looking at the highest selling car. What is the biggest seller for Toyota? the RAV4, which helpfully counts as "not a car".
I think for Toyota's main market the hybrid model still makes very much sense. EV is still a huge premium. Nice if you live outside the city and can charge your car with your own solar panels. Too clumsy to deal with in the city with the scarce public charge points.
And Toyota is a budget to mid range brand. For high end (the customers where the premium doesn't matter) they have lexus.
Massive amounts of anti-fossil fuel powered car legislation being passed, or being talked about, and Tesla being the clear market leader in the electric vehicle space.
I have no love for Tesla and I won't buy one as long as Elon Musk owns it.
But alas I think most major automakers are massively dropping the ball on EVs. I'm particularly disappointed in Toyota and Honda - I really expected better from them.
Where's the Corolla EV or the Civic EV? How about a nice little hatchback? Nope. Everyone's building crossovers, compact SUVs, fuckin' pickup trucks - it's a complete clownshow.
One thing that’s frustrating is that everyone is converging to a few form factors instead of trying to fill each other’s gaps. There aren’t really any sporty things, and even the F-150 electric is only barely at 150-class specs (the bed is tiny).
Also, the number of boneheaded moves by the legacy car companies is astounding.
Ford just pushed their roadmap out a few years. Stellantis has nothing(?). Chevy dropped the ball by discontinuing their best model and then deciding not to pursue the “people that own cell phones” market segment.
I think the list of top ten global automakers will be very different in ten years. These companies seem likely to be on it (no particular order), based on current lineups of native EV platforms:
Tesla, Rivian, Volvo (polestar), Kia, Mercedes, BMW, Hyundai, and probably Ford, despite the delays.
I’m leaving Genesis and Vinfast off my US-centric list, but they’re certainly contenders.
I suspect, without knowing the actual situation with Chevy, that this is a somewhat oblique reference to some grievance with their approach to Android Auto/CarPlay (or lack there of).
The extra up front cost and greater long term depreciation of a Corolla EV or civic EV is not worth it compared to a gas hybrid car for anyone that drives 10k miles per year, maybe even 15k miles per year.
Toyota’s gas hybrid tech has decades of track record that they will hold their value.
A Corolla or Civic EV implies a relatively cheap, really good quality, reliable car that holds its value. Not literally those bodies on an electric platform, but something that embodies those characteristics in an electric vehicle. I guess I used to have faith that if anyone could crack that problem, it would be the two Japanese giants.
The performance and cost of batteries is still changing fast enough that it doesn't make all that much sense for them to be dipping their toe in the water. They are investing in battery tech and are pretty likely to launch good vehicles when batteries that have good range and charge fast hit some target price point.
The market for electric vehicles that cost quite a bit more than similar size vehicles exists, but it's not like people that are shopping on a budget are going to choose based on performance and prestige, they are going to choose based on price, and conventional vehicles can still be built for lower costs.
The issue is, if you want to make an EV cheap enough to sit in the Civic/Carolla market, it still ends up being useless for most people because you can't afford to equip it with a large battery. Not to mention the only way an EV makes sense is if you can charge it with your own utility power, having to go and sit at a charger for 30+ minutes isn't, and will never be, tenable for most people.
Think about who's buying Civics and Corollas. It's apartment dwellers mostly, and they don't have chargers, nor can they. Maybe their parking lot has chargers, and those will be expensive, and they could use public chargers, which are also expensive, but you're talking basically making the cost 1:1 with a gas car at that point, but less convenient, less range, heavier, and all that assumes you can even sell it in that price bracket. It simply makes no sense for the majority of consumers. You'd have to invert the gas/electric pricing, which, given current trends, isn't likely to happen.
Now will that always be the case? Very likely not, eventually we'll have cheap power, eventually enough people will be driving EV's that gasoline loses its scale advantage. Until then, the EV market will stay mid-high end, and that means crossovers/suv's/trucks/etc.
Some people view that as dreadfully bad, myself, I'm happy that EV's are being developed and sold. By the time the world is ready for them, they're going to be very, very good, and that's the point which I'll be buying one. That is assuming I don't nab myself a late 90's ford ranger with a blown motor and build my own EV first.
Americans in the majority prefer pickup trucks, SUVs and crossovers. Ford entirely quit making sedans and hatchbacks a few years ago, at least for North America.
Tesla primarily sells into coastal cities where the driving conditions and use cases are ideal for an EV sedan. That said, most of the people I know with a Tesla also have an ICE SUV or truck.
The world is a lot bigger than the US. And valuation is not actual value but more whatever the current overhyped investor thinks it's going to bring in the future.
Tesla combines Model 3 and Model Y sales when they report results, so we don't actually know that. According to estimates*, and my own eyes, more Model 3s are sold than Model Ys.
* 2021 is the last year for which Wikipedia has estimates for both vehicles. The Model 3 sold over 1m, the Model Y around 400k.
Right but some countries (incl China) have explicit registration data so these estimates are not a shot in the dark. From Jose Pontes' article where the Wikipedia Y estimates came from, there were 530k 3s sold and 1.2M Ys sold
I owned a Tesla and could write a Master's dissertation on it. Interestingly, me buying the Tesla car was my life's shittiest investment but the Tesla shares I bought around that time (mid 2019) were my best investment.
I drove a Volvo after that, until landing in a very dangerous accident recently (the Volvo literally saved my life). Now I'm driving a rental Honda (which was my starter car on which I learnt to drive). Both beat Tesla by a mile.
Yeah Volvo is great at safety. And reliability too. I owned a 10 year old one that cost me a couple thousand, drove it for years without notable maintenance except the usual oil change and it kept going <3
So much better than my Skoda where half the dashboard fell apart and the gearbox suddenly just "crashed"
If I ever own a car again it'll be a Volvo :) won't be an EV for sure because I always buy old cars, i just hate spending a lot on a car. It's not important in my life and the past 6 years I haven't even owned one and rented one only once.
In my experience driving them. I like manual buttons and safety. Granted, the Tesla was the first car I bought and it was an amazing car in and of itself, but the ICE experience was far better. The software side sucks (I don't know why people keep praising Tesla for it).
Valuation is based on future, not past performance. Toyota is looking a bit stale and weak lately. They are loosing market share. Amongst others to BYD and Tesla in a market that Tesla single handedly created. And Toyota's portion of that is so pathetic it's barely worth talking about.
It's gotten so bad that Toyota is currently actually outsourcing EVs to BYD at the same time that company is grabbing market share from them. Not a great look if you have to ask your fastest growing competitor to please build a product for you.
Tesla is of course taking a few big bets that may or may not pay off. Many people are rightfully somewhat skeptical about all that but it is a bit disingenuous to portray them as just a car company given their growing number of "definitely not a car" type products and services. E.g. battery storage is bringing in quite a bit of revenue now for them and is one of the things that is rapidly growing. And of course the whole robots, AI, and FSD angle has lots of potential. Regardless of whether you believe Tesla can do that or not.
This article is about OpenAI who is getting quite stellar valuations in that area. Tesla is of course part of that market. Toyota very much isn't.
In short, Tesla might fail or they might succeed. In the same way Toyota might recover or they might not. To be honest, both companies have issues. But I have more confidence in Tesla than Toyota. Toyota seems a company without ambition and vision. Tesla can't be accused of that. Toyota's vision is to do what all their competitors are doing right now but in ten years. Tesla's vision is to do in ten years what nobody is doing right now.
Again, you might not believe in any of that. But you might be wrong. That's the gamble you make in the stock market with either company. And the share price is just a reflection of what people buying shares right now believe.
> The market correctly anticipated Tesla's popularity.
It seems like the market has priced in gains that have yet to materialize in Tesla's fundamentals. They're still far smaller than auto makers with much lower valuations.
I’d guess it’s some contracts that’s valueated at high multiple for all kind of reasons, like it’s enterprise deal so it’ll be there forever and only increase the amount; the apple deal probably was known back then but only public recently; the influential type of customer that could help them break into particular industries;
If you're going to make a 4 letter comment, at least be correct. Tesla in no way could be thought of as "short-term popularity." Every year, they've had increasing sales, about +20% between 2022 to 2023, and their latest quarter is still the highest sales of any quarter, on track to make $100 billion annual.
Yet, as mentioned elsewhere in this same comment chain, they're almost certainly overvalued and in a bubble. Not only that, but their overall image has suffered over the last few years thanks to Musk. There are plenty of concerns over their inability to produce a product of actual high quality that doesn't consistently fall apart, and their product line is stagnating without any new cars (cars, not Cybertrucks).
It's cool that sales have increased, but so many other signs point to their run being relatively short-lived that it's not an unreasonable viewpoint to expect them not to be as strong as they are long-term.
Edit: Hell, even Musk has said they're not a car company anymore. Can anyone actually tell me what kind of company they are, then? How can I expect for them to be as successful if they're having some kind of existential crisis, for lack of a better word? Good luck getting your employees moving towards the same mission if we're not entirely clear what that is.
Edit 2: Further, they've only been selling their current lines of cars for twelve years. That's a relatively tiny amount of time compared to other leading manufacturers. That's short, and all signs point to them not really knowing where to go from here. Just because there's been growth doesn't mean it's guaranteed to continue. Downvote this comment too, disagree with it, I don't really care - time will tell how this shakes out.
This scheme will soon become unsustainable like the last time the tech industry corrected in 2022.
Hopefully the participants don't experience any down-rounds or devaluations along the way. (It happened to many over-valued startups like Klarna, and Stripe and OpenSea.)