I no AI fanboy at all. I think it there won’t be AGI anytime soon.
However, it’s beyond my comprehension how anyone would think that we will see a decline in demand growth for compute.
AI will conquer the world like software or the smartphone did. It’ll get implemented everywhere, more people will use it. We’re super early in the penetration so far.
At this point computation is in essence commodity. And commodities have demand cycles. If other economic factors slowdown or companies go out of business they stop using compute or start less new products that use compute. Thus it is entirely realistic to me that demand for compute might go down. Or that we are just now over provisioning compute in short or medium term.
I wonder, is the quality of AI answers going up over time or not? Last weekend I spent a lot of time with Preplexity trying to understand why my SeqTrack device didn't do what I wanted it to do and seems Perplexity had a wrong idea of how the buttons on the device are laid out, so it gave me wrong or confusing answers. I spent literally hours trying to feed it different prompts to get an answer that would solve my problem.
If it had given me the right easy to understand answer right away I would have spent 2 minutes of both MY time and ITS time. My point is if AI will improve we will need less of it, to get our questions answered. Or, perhaps AI usage goes up if it improves its answers?
The problem is it's inability to say "I don't know". As soon as you reach the limits of the models knowledge it will readily start fabricating answers.
Both true. Perplexity knows a lot about SeqTrack, I assume it has read the UserGuide. But some things it gets wrong, seems especially things it should understand by looking at the pictures.
I'm just wondering if there's a clear path for it to improve and on what time-table. The fact that it does not tell you when it is "unsure" of course makes things worse for users. (It is never unsure).
Always worth trying a different model, especially if you’re using a free one. I wouldn’t take one data point to seriously either.
The data is very strongly showing the quality of AI answers is rapidly improving. If you want a good example, check out the sixty symbols video by Brady Haran, where they revisited getting AI to answer a quantum physics exam after trying the same thing 3 years ago. The improvement is IMMENSE and unavoidable.
With vision models (SOTA models like Gemini and ChatGPT can do this), you can take a picture/screenshot of the button layout, upload it, and have it work from that. Feeding it current documentation (eg a pdf of a user manual) helps too.
Referencing outdated documentation or straight up hallucinating answers is still an issue. It is getting better with each model release though
More so I meant to think of oil, copper and now silver. All follow demand for the price. All have had varying prices at different times. Compute should not really be that different.
But yes. Cisco's value dropped when there was not same amount to spend on networking gear. Nvidia's value will drop as there is not same amount of spend on their gear.
Other impacted players in actual economic downturn could be Amazon with AWS, MS with Azure. And even more so those now betting on AI computing. At least general purpose computing can run web servers.
Even suggesting that computers will replace human brains brings up a moral and ethical question. If the computer is just as smart as a person, then we need to potentially consider that the computer has rights.
As far as AI conquering the world. It needs a "killer app". I don't think we'll really see that until AR glasses that happen to include AI. If it can have context about your day, take action on your behalf, and have the same battery life as a smartphone...
I don’t see this as fanaticism at all. No one could predict a billion people mindlessly scrolling tiktok in 2007. This is going to happen again, only 10x. Faster and more addictive, with content generated on the fly to be so addictive, you won’t be able to look away.
What if its penetration ends up being on the same level as modern crypto? Average person doesn't seem to particularly care about meme coins or bitcoin - it is not being actively used in day to day setting, there's no signs of this status improving.
Doesn't mean that crypto is not being used, of course. Plenty of people do use things like USDT, gamble on bitcoin or try to scam people with new meme coins, but this is far from what crypto enthusiasts and NFT moguls promised us in their feverish posts back in the middle of 2010s.
So imagine that AI is here to stay, but the absolutely unhinged hype train will slow down and we will settle in some kind of equilibrium of practical use.
I have still been unable to see how folks connect AI to Crypto. Crypto never connected with real use cases. There are some edge cases and people do use it but there is not a core use.
AI is different and businesses are already using it a lot. Of course there is hype, it’s not doing all the things the talking heads said but it does not mean immense value is not being generated.
It's an analogy, it doesn't have to map 1:1 to AI. The point is that current situation around AI looks kind of similar to the situation and level of hype around Crypto when it was still growing: all the "ledger" startups, promises of decentralization, NFTs in video games and so on. We are somewhere around that point when it comes to AI.
No it’s an absolutely ridiculous comparison that people continue to make even though AI has well past the usefulness of crypto and at an alarming rate of speed. AI has unlocked so many projects my team would never have tackled before.
Anecdotally, many non-technical users or "regular joes" as it were that I know who were very enthusiastic about AI a year ago are now disengaging. With the rate really picking up the last couple of months.
Their usage has declined primarily with OpenAI and Gemini tools, no one has mentioned Anthropic based models but I don't think normies know they exist honestly.
The disengagement seems to be that with enough time and real world application, the shortcomings have become more noticable and the patience they once had for incorrect or unreliable output has effectively evaporated. In cases, to the point where its starting to outweigh any gains they get.
Not all of the normies I know to be fair, but a surprising amount given the strange period of quiet inbetween "This is amazing!" and "eh, its not as good as I thought it was at first."
However, it’s beyond my comprehension how anyone would think that we will see a decline in demand growth for compute.
AI will conquer the world like software or the smartphone did. It’ll get implemented everywhere, more people will use it. We’re super early in the penetration so far.