Peteragain

6 ms·

31.
by Peteragain 10mo ago
There is an argument, perhaps no longer PC, that the indigenous population used fire to hunt, and so burnt off regularly. Fires these days are indeed devastating because we try to stop them. Established eucalyptus trees also thrive after
32.
by Peteragain 10mo ago
I caved in South Australia many years ago and what is not obvious is that south Australian caves are predominantly horizontal and dry. I now live in the Peak District in the UK and caving is a muddy wet affair with too many ropes.
33.
by Peteragain 10mo ago
"they can articulate concepts.. that [don't exist] in the training corpus" yes, but that doesn't necessarily mean they have a model [of the world]. You might want to say they are articulating the plausible (that is somet
34.
by Peteragain 10mo ago
This is the problem. I am arguing there are no distributed representations (cf Harnad's original symbol grounding problem paper, Hinton, and others). There are "distributional representations" by definition (cf the Wikipedia
35.
by Peteragain 10mo ago
Thanks for the reference and I stand corrected. And yes I had looked at it a long time ago and will give it another read. But I think it is saying that RNNs are a means of approximating a statistical property of a collection of text. That p
36.
by Peteragain 10mo ago
Sorry for the late response. Yes that is Hinton's argument, and the claim made by the believers. On the other hand, if the GAC explanation is correct, an explanation might be that what we humans write down (that is, the training corpus
37.
by Peteragain 10mo ago
So, are current LLMs better because artificial neural networks are better predictors than Markov models, or because of the scale of the training data? Just putting it out there..
38.
by Peteragain 10mo ago
I must admit to having had an interest in AI for more than 20 years, and an attraction to museums and their nature for at least 10. I read this article from beginning to end and think it is really interesting. The idea that entertainment ge
39.
by Peteragain 10mo ago
Well yes.. the computer (actually the operating system) make it really hard to generate PWM and pigpio sorts that. The computer with an operating system gives you internet connectivity, ssh, power management...
40.
by Peteragain 10mo ago
The article is actually about the way we humans are extremely charitable when it comes to ascribing a ToM (theory of mind) and goes on to the Gym model of value. Nice. The comments drop back into the debate I originally saw Hinton describe
41.
by Peteragain 10mo ago
More expensive lawyers.
42.
by Peteragain 10mo ago
..and for his next trick, adding a servo is easy. Pigpio provides services underneath the operating system including square wave generators. Power the servo at 7.2 volts and drive the signal directly off a gpio pin at 5 volts!
43.
by Peteragain 10mo ago
The Flemming/Florey story is actually an interesting bit of science history, but it's about popular media and the romance of science v.s. how real science is done with patience, selfless dedication and a literature search. When Ch
44.
by Peteragain 10mo ago
The notion of "thinking" is not clear. I'll agree thinking with symbols is powerful and something (adult) humans and computers can do, but is is not the only way of making decisions. I'm going to suggest LLMs are not th
45.
by Peteragain 10mo ago
Cool!
46.
by Peteragain 10mo ago
That's how it works, but what does an LLM do is the real question. I'm working on the idea that this statistical model can be used for control. And that is enough for the evolution of agency. The claim from Vygotsky is that thin
47.
by Peteragain 10mo ago
One can certainly use language to _do_ things without thinking. Polly was a robot that gave a tour of the MIT labs, but it used pre recorded descriptions at various locations. The HUMANS gave meaning to the sounds.
48.
by Peteragain 10mo ago
I think the argument is that yes LLMs find patterns in token sequences. Assign tokens to moves in a chess game and the tokens are predictive of what happened in the past and of what chess players will do in the future. The LLM is not doing
49.
by Peteragain 10mo ago
Okay so rephrasing the question, how should we characterise the type of thinking we do without language? And the more interesting question IMO what thinking can an agent do without symbolic representation? The original Vygotsky claim was th
50.
by Peteragain 10mo ago
A beautifully written paper but I do feel it missed a major point. Vygotsky pointed out that "in ontogenesis one can discern a pre intellectual stage in the development of speech, and a pre linguistic stage in the development of though
51.
by Peteragain 11mo ago
Yes. The challenge is to understand how the river comes about, and at the same time, preventing it being gamed by vested interests.
52.
by Peteragain 11mo ago
Thanks for sharing. I think that, today, Ross is wrong and there is another way.. much like (I think) you are describing.
53.
by Peteragain 11mo ago
I am always delighted when I come across a good one, but finding good ones is the problem. Ross Wilkinson (CSIRO) many years ago said there were 3 ways to find things: browsing - from were you are, look about. If you're in the high str
54.
by Peteragain 11mo ago
I don't think there is a single point there that is not an assertion. You should become a BBC reporter.. although their assertions are usually reported as "xxx said..."
55.
by Peteragain 11mo ago
"How to steal a million" - a boomerang rather than to screwdriver..
56.
by Peteragain 11mo ago
Well there is a bit of history there. What you say is probably true, but I believe the British, then the the UN, and now a vocal majority, want exactly what you are talking about. My Jewish friends are happy with their Arab neighbours, but
57.
by Peteragain 11mo ago
Wow! Thanks! It was a half a thought but that interaction is right up there with "the big red button" and makes the last 20 years of enshitification all worth while!
58.
by Peteragain 11mo ago
And you are talking of current polling. I believe you will find that in Corbyn's hayday the polling was suggesting he would sweep labour to power.. but with the wrong politics. Hence being pro Palestinian was turned (in the bbc reporti
59.
by Peteragain 11mo ago
The BBC gets to say what history is because it documents the now. Farage does not have the popular support tha Corbyn had (has?) but you'd never believe it counting the number of interviews Farage gets with the BeeB. The game, as with
60.
by Peteragain 11mo ago
Okay they're dead, but I think the interesting thing here is the relationship between hardware and the way mathematicians (potentially) think about problem solving. The established practices massively constrain the solutions we find, b

More ›