All along I had thought that "AGI", "RSI", etc. were at the model level: but this paper seems to be talking about "agents", etc. I'm not sure having a swarm of agents explore a problem space in parallel via brute force is what "AGI" is about. I'd be happy to be proven wrong.
AGI and RSI are both meaningless terms, meaning whatever you choose them to mean.
RSI is the new sexy. Models are RSI-ing themselves towards the singularity, these folks' agents are RSI-ing themselves towards mastery of their training environments, and my pet cat is RSI-ing himself into the best cat that he can be.
> Ask yourself, how does Google -- a company that famously does everything -- benefit from SWEs outside of Google having access to powerful coding models?
Catch is, even Googlers internally do not have access to top-tier models. (or did not until recently, when apparently Claude was made accessible to the SWEs internally).
I don't understand the point of a "one time" tax. What happens once the money runs out a year later? Another "one time" tax?? Why not implement a permanent tax if you want to permanently solve the problem??
If they are blocking the one time tax, would they allow for a perma one. Don't be naive. No matter what type of tax it is, they will throw money to keep the status quo. $100M is chomp change to him and he can deduct it most likely.
It's basically to send a message to a group of people that they should fuck off. There is a reasonable gamut of opinions about whether this is valid. But I think it's fair to say Californians are upset about billionaires, and that this is a coherent way to message that opposition. (Similar to the dumbfuck states messaging around wokeness or whatever. A narrowing minority of America is prosperity first.)
Agreed. And his doom words have set a 1000 mouths in the Pentagon/Whitehall/August 1st Building/Kremlin salivating with excitement.
Take China, for example. Look at any recent ML conference, and see the fraction of articles majority-authored from Chinese universities and labs. Do you think they'll slow things down anytime soon? I don't think so!
It's a global arms race, and we're just spectators.
Yes, skills that are actually used count toward token consumption.
The question is whether the number of tokens required to achieve a certain behavior/intelligence/quality is equal between you manually providing those tokens via skills versus the model "deriving" the "skills" it needs on-the-fly in order to produce the outcome you want.
The claim above is that the former requires far fewer tokens.
Also skills only consume tokens when they are used, and part of the value is that the model will dynamically find and disclose only what's needed (assuming the skill is "well-designed").
Their claim is not about the prompt or skill tokens, it's about output tokens - skills can help the model bypass some thinking tokens or avoid reasoning deadends, and that way reduce output token usage. That's what they seem to have found empirically from their testing. (If it's truly 2x-4x, the time savings in waiting for the output is a pretty nice benefit too.)
Sergey Brin (of Google fame) was known for "dating" young interns/employees at Google in the early days. Once, when HR reminded him that he could get into trouble (since it was against policy), his response was, basically (paraphrasing) along the lines of "Why am I hiring them then?"
Basically, he considered it is his right to have his way with his young interns and employees.
I don't necessarily doubt that this happened, but it's worth pointing out that this anecdote is from Adam Fisher's 2018 report of what Charlie Ayers recounted that HR told him Brin said, twenty years after the fact. Maybe it happened that way, maybe not.
reply