this is literally how i got my first internship — it’s called google foobar
@jxmnop
-
What We Still Don’t Know About Language Models
By
–
people keep saying AI is moving so fast. some days I agree, but some days I'm not sure – so many papers published, but I don't feel like we're making that many fundamental breakthroughs. to cap off 2023, here's a list of things we still don't know about language models: – how
-
Language Models Power Compared to Nuclear and Internet Technology
By
–
I don't think language models are "the most powerful technology"; I think that title would go to something else, perhaps nuclear bombs, or battleships, or The Internet
-
Utilitarianism Without Proper Future Discounting in AI Ethics
By
–
I have long wondered if this is exactly the ideology that comes out when you’re purely utilitarian and fail to properly add a discount factor to valuations of future states
-
Why AI Alignment Undervalued Compared to Global Priorities
By
–
certainly AI alignment (on a high level) is an important issue but why is it undervalued relative to eg world hunger or pandemic preparedness? and given this, why is studying language models considered a reasonable path towards eventually saving the world?
-
Mechanistic Interpretability: Noble But Disconnected from AI Safety
By
–
I also think the sub field of mechanistic interpretability is very cool — it’s all three noble, challenging, and interesting — I just struggle to see how it connects to the broader goals (building AI systems that don’t kill us) or at least why it’s a top priority
-
Effective altruism goals versus narrow AI research priorities
By
–
i’m curious about effective altruism: how do so many smart people with the goal “do good for the world” wind up with the subgoal “analyze the neurons of GPT-2 small” or something similar?
-
Training Models to Generate Gist Tokens for Intermediate Reasoning
By
–
yeah this is very related idea — dense vectors convey more information than discrete tokens, in less space! but it's not the same thing, i'm describing how to train a model that generates gist tokens on the fly for its intermediate reasoning steps
-
Interpretability vs Efficiency Trade-offs in AI Models
By
–
ok good point good point i'm sacrificing all hope of interpretability for the sake of efficiency / performance latent scratchpad wouldn't be interpretable I think I've read chain-of-thought can also be misleading/wrong though even if it produces the right answer
-
Feedback Transformer: Angela Fan’s Expensive AI Architecture
By
–
the most similar thing i know of is the Feedback Transfomer from Angela Fan: https://
arxiv.org/abs/2002.09402 it's so expensive to run though