361–370 of 443 posts
Earlier quoted context omitted.
Every cafe, airport, school I've been to has people using ChatGPT or its competitors. Its obviously valuable for almost anyone. Just like how people cant imagine life before smartphones, people wont be able to imagine life before LLMs became ubiquitous. Its everywhere.
I’m sorry, but what? How can you tell people at the cafe/airport/school are using LLMs?
Earlier quoted context omitted.
The Nissan Leaf was far more affordable than a Tesla 10 years ago and very practical for anyone living in a city.
Eh it was pretty limited. The Leaf (then) couldn't go from my house, to the airport in my city (Melbourne) and back on one charge. That always made it a dealbreaker for me. And that's going by Nissan's claimed range, not even real world. So that's on a 100% charge, when the car is brand new with no battery degradation, and under the ideal efficiency conditions that you never really get.
Earlier quoted context omitted.
The Nissan Leaf was far more affordable than a Tesla 10 years ago and very practical for anyone living in a city.
While it was an affordable vehicle, saying that it was practical is an overstatement. Charging networks were abysmal and actually still are for non-Tesla compatible vehicles. If you had experience using EVgo and similar small networks you probably wouldn't sound as confident.
Earlier quoted context omitted.
Google is a digital advertising company. OpenAI hasn't even entered the ads business. In the absolute best case they can take over a large chunk of Google's search market share, sure, but that still doesn't make it anything similar to Google in terms of finances. How do they start making the queries profitable? What do they do when their competitors (Claude, Gemini, Llama, Mistral, Grok and several others) undercut t…
Google also offered a free product. OpenAI isn’t offering a free product but a subscription product plus metered API product amongst other. Their economics are structurally better than googles assuming they can keep growing their captured market share. Their outrageous costs are also opportunities to optimize costs, including massive amounts of R&D, etc. They don’t need to be profitable now - in fact as bezos demonst…
Are they? I would guess that the cost per query for Google, even back then was insignificant compared to how must OpenAI is spending on GPU compute for every prompt. Are they even breaking even on the $20 subscriptions?
During their growth phase Google could make nothing from most of their users and still have very high gross margins.
OpenAI not only has to attract enough new users but to also ensure that they are bringing more revenue than they cost. Which isn’t really a problem Google or FB ever faced.
Of course presumably more optimized models and faster hardware might solve that longterm. However consumers expectations will likely keep increasing as well and OpenAI has a bunch of competitors willing to undercut them (e.g. they have to keep continuously spending enough money to stay ahead of “open/free” models and then there is Google who would probably prefer to cannibalize their search business themselves than let someone else do it).
> in fact as bezos demonstrated with Amazon for many years, profit is an indication you’ve run out of uses for capital to grow.
Was Amazon primarily funding that growth using their own revenue or cash from external investors? Because that makes a massive difference which makes both cases hardly comparable (Uber might be a better example).
I hope better competition appear before the enshittification begins. As far as I understand it they're actually underwater on their API and even $20/month pricing, so we'll either see prices aggressively increase and or additional revenue streams like ads or product placement in results. We've witnessed that every time a company's valuation is impossibly high: They do anything they can to improve outlook in an attemp…
> the enshittification I've assumed that when AI becomes much more mainstream we'll see multiple levels of services. The cheapest (free or cash strapped services) will implement several (hidden/opaque) ways to reduce the cost of answering a query by limiting the depth and breadth of its analysis. Not knowing any better you likely won't realize that a much more complete, thoroughly considered answer was even available…
This already happens. Many of the cheap API providers aggressively quantize the weights and KV cache without making clear that they do.
Earlier quoted context omitted.
They accomplished AGI (artificial general intelligence) years ago. What do you think ChatGPT is? Alternatively, what are you imagining this “AGI” you speak of to be?
They haven't. Why are you stating lies as facts?
General: able to solve problem instances drawn from arbitrary domains.
Intelligence: definitions vary, but the application of existing knowledge to the solution of posed problems works here.
Artificial. General. Intelligence. AGI.
As in contrast to narrow intelligence, like AlphaGo or DeepBlue or air traffic control expert systems, ChatGPT is a general intelligence. It is an AGI.
What you are talking about is, I assume, a superintelligence (ASI). Bostrom is careful to distinguish these in his writing. Bostrom, Yudkowsky et al make some implicit assumptions that led them to believe that any AGI would very quickly lead to ASI. This is why, for example, Yudkowsky has a very public meltdown two years ago, declaring the sky is falling:
https://www.lesswrong.com/posts/j9Q8bRmwCgXRYAgcJ/miri-annou...
(Ignore the date. This was released on April 1st to give plausible deniability. It has since become clear this really represents his view.)
The sky is not falling. ChatGPT is artificial general intelligence, but it is not superintelligence. The theoretical model used by Bostrom et al to model AGI behavior does not match reality.
Your assumptions about AGI and superintelligence are almost certainly downstream from Bostrom and Yudkowsky. The model upon which those predictions were made has been falsified. I would recommend reconsidering your views and adjusting your expectations accordingly.
Earlier quoted context omitted.
Obviously, that's my point. We can do the math. GPT-4o can emit about 70 tokens a second. API pricing is $10/million for output tokens and $2.5/million for input tokens. Assuming a workload where inputs tokens are 10:1 with output tokens, and that I can generate continuous load (constantly generating tokens). I'll end up paying $210/day in API fees, or $76,650 in a year. Let's assume the hardware required to service…
Inference alone totally can be. Just look at banana.dev, runpod, lambda labs, or replicate. The issue is OpenAI is not just selling inference. Though I wouldn’t be surprised if there were some hidden costs that are hard for us to account for due to the sheer amount of traffic they must be getting on an hourly basis.
Earlier quoted context omitted.
They accomplished AGI (artificial general intelligence) years ago. What do you think ChatGPT is? Alternatively, what are you imagining this “AGI” you speak of to be?
OpenAI defines AGI as "autonomous systems that outperform humans at most economically valuable work." ChatGPT is not autonomous or capable of doubling global GDP.
The founders of OpenAI were drawn from an intellectual movement that made very specific, falsifiable predictions about the pipeline from AGI (original definition) to superintelligence, predictions which have since been entirely falsified. OpenAI talks about AGI as if it were ASI, because in their minds AGI inevitably leads to ASI in very short order (weeks or months was the standard assumption). That has proven not to be the case.
Earlier quoted context omitted.
> Google's current market capitalization is in the trillions. 2 trillion. Approximately 13x OpenAI's current valuation. Google nets almost 100 billion a year. OpenAI grosses 4 billion a year. Wild numbers.
A private pre-IPO investment is a bet on where OpenAI will be 10 years from now, not where they are now.
Earlier quoted context omitted.
you don't need a moat during the gold rush. You need scale - largest number of biggest shovels with which the stuff to be shoveled the fastest. There is so much money right now to be sucked up from the world. We're talking valuations in AI at $100M+ per employee. https://finance.yahoo.com/news/uae-backs-sam-altman-idea-095... >The models will have diminishing returns Wasn't that the going thinking before ChatGPT? And…
> the stuff to be shoveled They are spending a lot on shovels but it’s not clear that there is that much “stuff” (consumer demand) to be shoveled. VC money can only take you so far, you still need to have an actual way of making money. LLMs might effectively replace Google but they are already a commodity. It’s really not clear what moat OpenAI can build when there are already a bunch of proprietary/open models that…