Searching...
14 results for “proximity principle”
It's because I think the arc of intelligence has to go to what Justin calls affordances. And the arc of intelligence, if you look at evolution, right, the arc of intelligence eventually enables animals and humans, especially human as an intelligent a
29:14 / 30:20
from transcript29:14 – 30:20
...its proximity. I don't think that itself is particularly weird. I really like this term, the Don Rowe doctrine, because there's two things. There's this long standing history that The US wants to have a role to play in everyone else's politics among
5:42 / 7:01
from transcript5:42 – 7:01
...and principle is essentially he's the president, and what's in his head and his ideas are legitimate because they're his ideas and he's the president. So, yes, there's precedent.
5:14 / 6:15
from transcript5:14 – 6:15
...in proximity to this other phone because they have spyware on it. And now you know exactly when this person goes and plays golf. You know where they're gonna be on the golf course, and boom, just like that lunatic, the second person who was trying to
52:21 / 53:39
from transcript52:21 – 53:39
there's some something rustling in that bush. Don't go near. We know how to react to that data. We know how to, save ourselves. We internalize that learning, and our brain cells or our synapses remain plastic throughout our lifetime. What happens wit
24:01 / 25:00
from transcript24:01 – 25:00
You just pass this to the to the LLM as a prompt, and it processes it exactly the way it would process any other prompt, which is not an example of in context learning. So that really means that the underlying mechanism is the same. Right? Whether yo
26:56 / 28:01
from transcript26:56 – 28:01
that all of the models learn very similar things. Like, we have some preliminary evidence for that. Like, certain models that are fine tuned from the same base, you can kind of swap their representations without doing much, or if you look at the near
32:58 / 34:08
from transcript32:58 – 34:08
So that's an example where it has seen DSLs or structures in the past. And now using this evidence that I show, okay, this is what my DSL looks like. Now a new natural language query, it is able to create the right posterior distribution for the toke
26:12 / 27:17
from transcript26:12 – 27:17
And the scale word gets a lot of attention in this. The interpretation that I use is effectively to avoid adding in the human priors to your learning process. And if you read the original essay, this is what it talks about is how researchers will try
50:13 / 51:23
from transcript50:13 – 51:23
And therefore, it's trying multiple solutions, and it it emerges this chain of thought. This might be a good place to, to mention the, the eloquent and the insightful tweet of the great and the powerful Andre Karpathy. I think he had a bunch of thoug
2:54:42 / 2:55:46
from transcript2:54:42 – 2:55:46
it'll cause this much pain or all that. Correct. What you're essentially doing in your head is you're doing a simulation. You see the, the the the the pen coming, and you know that it'll come and hit me. Your mind simulates, and you dodge it. Right?
27:43 / 28:49
from transcript27:43 – 28:49
you essentially have seen the whole Internet. So you have definitely seen people explaining their work even even verbally, like a transcript of a math lecture. You try this, oh, I messed this up. And what reinforced learning is this RLVR is very good
1:58:22 / 1:59:42
from transcript1:58:22 – 1:59:42
This has been an episode of Unsupervised Learning, an AI podcast by Redpoint Ventures, where we probe the sharpest minds in AI about what's real today, what's going to be real in the future, and what it means for businesses in the world. With the fas
1:01:20 / 1:01:44
from transcript1:01:20 – 1:01:44
And an interesting kind of technical part of this is that, I mentioned what we really care about is this, how much new information do you learn from the next data point? So technically, that's the marginal information gain per data point. Perplexity
27:40 / 28:50
from transcript27:40 – 28:50
Have a podcast?
Run your podcast through Clypt
Get ranked clips, hooks, and ready-to-post copy from your own episodes. Free to try.