Vadim Liventsev

llm research, llm wrappers, llm rappers

[Vadim Liventsev @vadimdotme]

guys only want one thing and it's a badge that says "member of technical staff"

[Vadim Liventsev @vadimdotme]

gonna train some custom models to map

the problem + Claude/GPTs opinion on the problem -> solution

could be very powerful for small specialised datasets (or mode collapse into just parroting the oracle)

[Vadim Liventsev @vadimdotme]

@eterevsky @peterrhague in addition to the cooling effect of solar panels themselves!

[Vadim Liventsev @vadimdotme]

sweating rn because USG put export controls on these insights

[Patrick Collison @patrickc]

I asked Claude about the air conditioning debate in Europe, and it really didn’t pull any punches.

[Vadim Liventsev @vadimdotme]

@broblas @paulg identity politics is much worse in the UK

[Vadim Liventsev @vadimdotme]

@dionizije_fa some of us do pushups at home

[Vadim Liventsev @vadimdotme]

every room on the planet should have a thermostat that works from +15C to +25C

whether it requires AC or not is an implementation detail, in some parts of the world it doesn't, but keeping people in rooms without one will be considered a form of torture by our grandchildren

[Vadim Liventsev @vadimdotme]

@xav_db your regular reminder to donate to @MakeSunsets

[Vadim Liventsev @vadimdotme]

your regular reminder to donate to @MakeSunsets

[Vadim Liventsev @vadimdotme]

this is what the "flights to anywhere" search engines very invented for

[Vadim Liventsev @vadimdotme]

this is what the "flights to anywhere" search engines very invented for

[Vadim Liventsev @vadimdotme]

@benlandautaylor Netherlands mentioned

[Vadim Liventsev @vadimdotme]

do we need a pause on advancing the capabilities of the AI alignment community until we make sure the AI alignment community is aligned with the interests of humanity

[Vadim Liventsev @vadimdotme]

@AmandaAskell the doctors have been trained to minimise “sycophancy”

[Vadim Liventsev @vadimdotme]

@segyges not offending a very specific subset of customers

[Vadim Liventsev @vadimdotme]

as a result, the talent pool of people most qualified to build powerful AI consists almost exclusively of people that in early 2000s got scared of powerful AI killing everyone

[Vadim Liventsev @vadimdotme]

there's a more pragmatic explanation: mechanistic interpretability, widely known in AI safety circles as an alignment technology, turned out to be an outstanding capabilities technology

[Richard Ngo @RichardMCNgo]

The AI safety community constructed a memeplex in which “taking AGI seriously” was a prerequisite for being a serious and good person. When inside this memeplex (as many at Anthropic, some at OpenAI, and a few at DeepMind are) your vision narrows until the world feels extremely constrained. The whole future seems to flow through the “one ring” of controlling recursive self-improvement. And so even when you worry about AI itself seizing that one ring, you can’t generate better strategies than trying to control it yourself (directly via an AGI company, or indirectly via AGI governance).

I’m not saying this is a pure hyperstition. There’s a core truth underlying this perspective: AI will become extremely intelligent and capable, much more than it is today. But the current world is much more spacious and human-empowering than the future which Eliezer originally envisioned (a “brain in a box in a basement” taking over the world by surprise). And it would be even more spacious if this memeplex weren’t active. For example, Satya and Mark and Sundar only started taking AGI seriously because OpenAI forced them to—and even now they don’t really believe in superintelligence—and even if they did they couldn’t get most of their employees on board. Imagine how chill a “race” between Microsoft and Meta and Google would have been, compared with what we have today: Dario and Sam deep in the “one ring” memeplex while also personally loathing each other.

So the one ring memeplex has an escalating life-cycle. It infects people by letting them harness the narrative that they’re good people for taking AGI seriously, and that making other people take AGI seriously is a boon for the world (despite how terribly that’s gone so far). Then it shuts off their imagination—any sparks of creativity or plans that don’t steer towards the one ring are quickly shut down. Instead they make ChatGPT or the METR graph or other recruiting tools for the memeplex. And yes, they’ll acknowledge that previous versions of the memeplex were too extreme, and led to overly constricted action. But we don’t have time to worry about that, they’ll say, because AGI is coming by 2027/2028, and that’s the end of history. Somehow, though, almost everyone with that view has only a vibes-based definition of AGI. They don’t believe in Dyson spheres by 2028, or self-replicating nanotech by 2028, or brain emulations by 2028. They mostly can’t make concrete predictions, except that it’ll be enough AI that it puts all their plans on a deadline. (Shout-out to @DKokotajlo and @paulfchristiano though, who do make concrete predictions about things going crazy soon.)

It seems very hard to break out of this memeplex without just giving up. David Holz is maybe the world champion of that—the only person who was in a position to race for AGI and consciously turned away. Various agent foundations researchers have carved out space to think real thoughts, not the kind of panicky stabbing in the dark that usually passes for safety research. A few others (e.g. Salamon, Hoffman, Vassar, Andre, Sahil, Davidad) are pursuing more unusual paths. And of the people who burned out, I expect some will reorient to doing creative thinking.

For others, the main takeaway: yes, the future of AI will be wild. But so far it’s increased peak human agency, and openness to this trend continuing over the next decade will allow you to start creating something worth creating.

[Vadim Liventsev @vadimdotme]

@_xjdr and you have the best AI lab website I have seen to date :)

[Vadim Liventsev @vadimdotme]

@CAIS "Countries associated with repression, censorship, or opaque governance rank low"

Fable literally puts the UK at S-tier

[Vadim Liventsev @vadimdotme]

@redtachyon https://t.co/nVQvTRtPsP

[Vadim Liventsev @vadimdotme]

Chinese AI labs are Jeffersonian

American AI labs are Maoist

[Vadim Liventsev @vadimdotme]

@Laz4rz Anthropic folks didn't share this sentiment did they