“Not even wrong” …
AI 2027
201–210 of 641 posts
Re: AI 2027
#202Earlier quoted context omitted.
TBH, this kind of reads like the pedigrees of the former members of the OpenAI board. When the thing blew up, and people started to apply real scrutiny, it turned out that about half of them had no real experience in pretty much anything at all, except founding Foundations and instituting Institutes. A lot of people (like the Effective Altruism cult) seem to have made a career out of selling their Sci-Fi content as p…
c'mon man, you don't believe that, let's have a little less disingenuousness on the internet
How would you know what he believes?
There's hype and there's people calling bullshit. If you work from the assumption that the hype people are genuine, but the people calling bullshit can't be for real, that's how you get a bubble.
Re: AI 2027
#203This is worse than the mansplaining scene from Annie Hall.
You mean the part where he pulls out Marshal McLuhan to back him up in an argument? "You know nothing of my work..."
Re: AI 2027
#204The limiting factor is power, we can't build enough of it - certainly not enough by 2027. I don't really see this addressed.
Second to this, we can't just assume that progress will keep increasing. Most technologies have a 'S' curve and plateau once the quick and easy gains are captured. Pre-training is done. We can get further with RL but really only in certain domains that are solvable (math and to an extent coding). Other domains like law are extremely hard to even benchmark or grade without very slow and expensive human annotation.
Re: AI 2027
#205> "resist the temptation to get better ratings from gullible humans by hallucinating citations or faking task completion"
Everything this from this point on is pure fiction. An LLM can't get tempted or resist temptations, at best there's some local minimum in a gradient that it falls into. As opaque and black-box-y as they are, they're still deterministic machines. Anthropomorphisation tells you nothing useful about the computer, only the user.
Re: AI 2027
#206As someone who's fairly ignorant of how AI actually works at a low level, I feel incapable of assessing how realistic any of these projections are. But the "bad ending" was certainly chilling. That said, this snippet from the bad ending nearly made me spit my coffee out laughing: > There are even bioengineered human-like creatures (to humans what corgis are to wolves) sitting in office-like environments all day viewi…
Sigh. When you talk to these people their eugenics obsession always comes out eventually. Set a timer and wait for it.
Re: AI 2027
#207I think it's worth noting that all of the authors have financial or professional incentive to accelerate the AI hype bandwagon as much as possible.
I realise no one is infallible but do you not think Daniel Kokotajlo's integrity is now pretty well established with regard to those incentives?
Re: AI 2027
#208Ok, I'll bite. I predict that everything in this article is horse manure. AGI will not happen. LLMs will be tools, that can automate away stuff, like today and they will get slightly, or quite a bit better at it. That will be all. See you in two years, I'm excited what will be the truth.
I predict AGI will be solved 5 years after full self driving which itself is 1 year out (same as it has been for the past 10 years).
...not before I get in peak shape, six months from now.
Re: AI 2027
#209Readers should, charitably, interpret this as "the sequence of events which need to happen in order for OpenAI to justify the inflow of capital necessary to survive". Your daily vibe coding challenge: Get GPT-4o to output functional code which uses Google Vertex AI to generate a text embedding. If they can solve that one by July, then maybe we're on track for "curing all disease and aging, brain uploading, and coloni…
You’ve intentionally hamstrung your test by choosing an inferior model though.
Re: AI 2027
#210Much has been made in its article about autonomous agents ability to do research via browsing the web - the web is 90% garbage by weight (including articles on certain specialist topics). And it shows. When I used GPT's deep research to research the topic, it generated a shallow and largely incorrect summary of the issue, owning mostly to its inability to find quality material, instead it ended up going for places li…
This story isn’t really about agents browsing the web. It’s a fiction about a company that consumes all of the web and all other written material into a model that doesn’t need to browse the web. The agents in this story supersede the web.
But your point hits on one of the first cracks to show in this story: We already have companies consuming much of the web and training models on all of our books, but the reports they produce are of mixed quality.
The article tries to get around this by imagining models and training runs a couple orders of magnitude larger will simply appear in the near future and the output of those models will yield breakthroughs that accelerate the next rounds even faster.
Yet here we are struggling to build as much infrastructure as possible to squeeze incremental improvements out of the next generation of models.
This entire story relies on AI advancement accelerating faster in a self-reinforcing way in the coming couple of years.