That's interesting to hear. I should have added that I use Gemini through Google AI Studio as my general chat model, which probably explains our wildly different experiences.
I'm interested how to mitigate risk without over-regulating & stifling innovation from exactly the people who are most thoughtful about the risk (which is who we should want innovating & leading).
Even though this article is new, I'm not sure it has materially new ideas vs what some other labs have been claiming for years.
Enforcing safeguards that apply globally to good and bad actors alike so far as proved quite difficult - compute is essentially the only lever, and that's tenuous at best.
Unlike nuclear arms (where functionally there's a limit at MAD), so far we believe that with enough compute, energy, and algorithms, intelligence will continue to scale (and not only scale - but compound)
Do you still believe this considering those companies live and die by the CCP? Would you feel differently if scale was reversed and Meta was the upstart trying to compete with Xiaomi/etc?
I'm right there with you questioning Meta's value for society, but I do believe it has less opportunity for abuse/manipulation
Considering Meta lives and dies by the White House, where abuse and manipulation have become SOP, I'm not sure the CCP stacks up as the boogeyman anymore.
IMO opinions of this variety often come from being misinformed or under-exposed to exactly how oppressive the CCP is. The US government isn't perfect, but it values various freedoms much more extensively than the CCP
I don't care about internal politics of a state, as long as they are not killing people deliberately on a daily basis or invading other countries and killing their people.
How government is structured is up to Chinese people to decide
> Would you feel differently if scale was reversed and Meta was the upstart trying to compete with Xiaomi/etc?
It's not only about scale and upstart, it's also about how much are they harming society.
I've been showing people chatjimmy for months - it's incredible. Both reasoning and tool use generation scale with TPS. Imagine 100x more reasoning on a model, or 100x parallel tool uses.
I'll hijack this point to talk about robotics - I think we agree 1) It's not ideal for China to control production for _all_ of the things we use here.
2) What they already are good at they'll continue to improve at, because of domain knowledge & ecosystem
3) The reason they are ahead in the first place is labor cost & the QOL workers will tolerate (and intentional government investment)
4) The only way around this is for the US to invest into robotics (labor) & subsidize american manufacturing (ecosystem)
> The reason they are ahead in the first place is labor cost & the QOL workers will tolerate
That hasn't been accurate in a decade. Chinese labor costs have been higher than Mexico since before Covid. They have more efficient firms, business processes and so much more. The parenthesized "(and intentional government investment)" is related to how they developed nimbler systems, but ignores the totality of industrial policy.
I totally agree that we need to reinvigorate manufacturing in the US and that it’s going to take the government (federal level) to make this happen. But for some reason we’d rather give those tax dollars to billionaires and elect incompetents like the current president.
Without the singularity, I think Frontier labs will offer intelligent model blends. They'll have their own versions of "cheap" models and be expert at using the appropriate amount of compute for a task.
Frontier models need to do everything for everyone. It's expected (though not often done) that smaller models fine-tuned on specific tasks can approach frontier performance on a specific area. [0]
Post-training/fine-tuning is not trivial and having it as a service might make it more accessible.
In my experience everything that affects engineers affects an agent. Good abstractions, reasonably sized methods, good names, principled (intra & inter) service architecture, unit tests, etc.
All of these things have historically been the job of engineers, because it helps other people contribute to the code.
Now it helps other people and other agents contribute to the code.
It has taken LLMs to encourage companies to prioritise a clean codebase, tickets with unambiguous context and examples of what is right and wrong when onboarding new team members.
Related - "The Electric Slide" [0] from NotBoring (a techno-optimist news letter) talks about production of batteries and other parts of the "Electric Stack", and explains where the US/China are relative to each other and the rest of the world, and why China has such a big lead.
Not sure on consumer/product use though