The DeepMind Institute(institute.deepmind.com) |
The DeepMind Institute(institute.deepmind.com) |
- need for faster, better and more accurate measurements of the things we care about
- 3 scenarios of impact ranging from mild to major disruption
- mild policies are all sensible, like expanded unemployment insurance and Earned Income Tax Credit
- major disruption policies are also pretty logical, emphasizing owning a share of the profits of AI
- AI evaluators to sort and weigh the policies for their effectiveness
It looks like to me a lot of thinking, research, and work put into the piece.
I know people will gripe at using AI evaluators to score and weigh policies for effectiveness but if you read any amount of research papers, they do researcher evaluated scoring, which translates to "me, a 26 year old PHD candidate who hasn't held a job yet, and my roommate who stayed up all night with me to tell me his vibes on the subject".
It is IMO wrong and lazy to dismiss these pieces as being the same level as Dario or Sam's blog posts / speeches / call to actions, these are actual research papers.
Not in any sense of that term are these research papers. Not lazy, sure, but a couple conceptual tables does not a research paper make.
Consequently things like "Sovereign AI dividends" and the like won't matter in the extreme cases. Profits on a scale that matters for society AI itself will only be a thing in an intermediate scenario which will probably only be real for an incredibly short time.
And what about for the rest of the world? What do you do with all the people in middle income countries that don't have an economic and tax structure like the US?
This looks like a new account from a PR agency focusing on AI labs.
The fact that these come from bots at least gives me hope that the fact a few good posts come up every day from the sea of slop means the actual humans still appreciate good writing.
Today’s AI systems have impressive capabilities and the rapid pace of innovation suggests we’re now approaching artificial general intelligence (AGI), a system that exhibits all the cognitive capabilities of the human brain.
The best part of technology: the people who brought you the "disruption" are the ones who promise you they have the solution to all the problems it caused.
Google, who was left a little bit behind anyway, is more than happy to join the calls for taking a more cautious approach to AI development. Hence DeepMind Institute.
Sam Altman has no option than to pretend he agrees with all these ideas. What else is he supposed to say?
> DMI is a platform for researchers and thinkers from across Google DeepMind, Google, and the wider global research community to work on and publish creative, deeply informed ideas about a world with AGI. They will not always agree, and they will likely change their minds, as more data and information comes to light at the fast-moving frontier. We created this institute precisely because broad-based intellectual discussion and debate are required to arrive at a consensus about how to address the challenges and opportunities we face as a society.
https://institute.deepmind.com/essays/the-case-for-reasoning...
(Hiding chains of thought is one of the most harmful things I think OAI/Anthr. has done, altough Google Gemini also encrypts its chains of thought, so it's interesting to see the DeepMind Institute crticising this.)
Of course this still then has a 'trust me, bro' feel to it, but making it outright impossible to reasonably inspect the CoT is a significantly larger and riskier step, IMO.
> We created this institute precisely because broad-based intellectual discussion and debate are required to arrive at a consensus about how to address the challenges and opportunities we face as a society.
Something tells me that consensus will never arrive to the conclusion that we should shut down that whole industry
>"Someone once said that it is easier to imagine the end of the world than the end of capitalism."
They cite Marxist philosopher Fredric Jameson (and misspell his name, incidentally), but I've always attributed this phrase to Fisher. I guess citing kpunk would have probably made the simulation we all live in actually implode.
Ai separated humans from language, like the writing did for memory and the printing press for knowledge distribution.
"AI does to us what American Cheese did to food" by Opus 23
https://www.youtube.com/watch?v=1CbmC2aWhjY
(still the best Ai video I have seen, also a work of art)
In which capabilities are AI not already close to or beyond that of a human brain?
* Forming a long lasting memory from a single encounter
* Speed of adaptation to new environments
* Deft manual dexterity, like playing the violin and needlework
* Competing in a triathlon (requires running, biking, and swimming)
* Having a single agent that can do ALL of those things and more
Humans understand and perform tasks in many more than just the few modalities current model providers have focused on. I'm not discounting the utility of current models in the limited domains they operate in, but they are no where close to being generalists like a human or other animals.
Rap lyrics are a good test. They are on the level of a censored version of Insane Clown Posse. Just terrible.
None of the models I've downloaded have sprung to life spontaneously, or done anything remotely unpredictable. None of them can act without input. They all have strictly limited ways to receive input and pass output. They aren't alive, can't feel, don't have experiences. The fact that they are a phenomenon that emerge from billions of snippets of our language lets them express languages in extremely convincing ways, and they are extremely impressive in what they can do. But this isn't Short Circuit or D.A.R.Y.L. or the Terminator, none of them have feelings. None of them continue to think after the program exits.
Without those characteristics they can only generate based on what they've ingested. The capabilities are still astonishing, but they're still dead things, and nobody will ever risk their life to protect the model "context" from getting "killed"
Transformers can also become confused be texts that no humans become confused by. I wrote some stories that I've used as test material, where I deliberately refuse to say who is speaking, or whose perspective we see, but where it is obvious to a human who it must be, and LLMs can make huge screwups in those texts. Mixing up an old guy with a young guy who he, when he was a similar age, was similar to, mixing up a kid with the kid's mother, that sort of thing.
I think the first part has no know solution. The second part is probably solvable, but not with a transformer-- maybe if they could make notes or output reasoning traces during prefill.
Both problems are hard and unsolved though.
It can’t even play a basic video game!
Now it can
It can never beat the best human go player!
Now it can
It can’t even make a video of will smith eating food!
Now it can
It can’t do basic math!
Now it can
It can’t code as well as a junior?
Not it can
It can’t pass the bar or other certificate exams!
Now it can
It’s never going to meet senior level programmers!
Now Linus says it’s better at coding than him
It can’t do high level maths!
Now it can
It can’t solve math problems humans cant!
Now it can
It can’t do RSI
…
For a guy who spent his life imagining and realising dreams in video games and cognitive abilities in artificial neural networks, it has to be very soul destroying for Demis to spend his time doing something as frivolous as this.
Encrypting CoT is just plain nuts yet we’ve all accepted it as normal.
That old thing?
seems like a very short timeline to me.
It is the only video in that channel, and worth it.
He has some ambitions for more videos, based on comment replies, but it would be way more legendary to have a single video forever, make a new channel and I'll sub to that one too
This blob of text seems to mean nothing
Maybe you need to learn more about the language uplift.
"writing" isn't a technology in a video game that you research and improves the knowledge preservation instantly. It has a lot of components that it needs to be effective in the way that is described here, and it really wasn't that effective even coming to way later ages than the age that it was "found"(?).
The relations described here aren't anologous to "Ai separated humans from language" so they seem like just some arbitrary things to say on the side.
why is writing things down better?
For the same reason we make kids write out math problems rather than just reciting them. It engages different parts of the brain for one, you gain an expanded amount of context by doing so.
The next thing has to do with human memory, when you remember something it is never read only. Recalling memory can adjust your neurons and change the memory, sometimes a little, sometimes a lot. Hence why eyewitness testimony is mostly trash.
Writing it down also provides a larger workspace for making changes that you can loop over. Quite often things sound fine in our mind, then we write it, and reread it and we realize its wrong. These days people commonly use this in agentic workflows. One agent will conceptualize an idea, and then it's passed off to another agent with a clean context to review for logical flaws. Reading ones own text does the same things for humans, we don't load the full memory context of what we wrote and it seemingly allows us to iterate over the problem differently.
And I guess lastly, human voice is very low bandwidth. A wagon full of well written books is far easier to manage than the humans that contain that knowledge. The reader is at their leisure to learn and the speaker isn't inconvenienced by being at someones beck and call.
Writing locks in and requires the evolution of other forms of evolution, as the Word can no longer evolve during transmission.
I say that while being firmly on your side in the larger back-and-forth exchange here :)