851–860 of 1001 posts
Ronan Farrow here. Andrew Marantz and I spent 18 months on this investigation. Happy to answer questions about the reporting.
I know why the cantilevered pool statement is there and why you mentioned it. I’m sure you don’t know half of the totally fucked up things Sam did to get “revenge” for the slight of a leaking pool.
Earlier quoted context omitted.
I wouldn't expect any LLM to be able to respect such a request. Do they even have direct access to published works to use as reference material? Also, last time I played 20 questions with ChatGPT, it needed 97 turns and tons of my active hinting to get the answer.
>>Do they even have direct access to published works to use as reference material? I mean, clearly, given that it did answer my question eventually. Also wasn't it a whole thing that these models got trained on entire book libraries(without necessarily paying for that). >>I wouldn't expect any LLM to be able to respect such a request Why though? They seem to know everything about everything, why not this specifically…
The problem with this line of reasoning is that it is unscientific. "They seem to" is not good enough for an operational understanding of how LLMs work. The whole point of training is to forget details in order to form general capability, so it is not surprising if they forget things about books if the system deemed other properties as more important to remember.
Ronan Farrow, the write of this article, made a comment in this thread that is buried in all the comments, "As is always the case with incredibly precise and rigorously fact-checked reporting like this, where every word is chosen carefully (the initial closing meeting for this one was nearly eight hours long, with full deliberation about each sentence), there is more out there on that subject than is explicitly on th…
I don't get why writers do this. It makes sense for fiction, but why a factual, non-fiction article?
I am just speculating but if @ronanfarrow is still checking the discussion here, it would be amazing to hear the actual reasons.
Earlier quoted context omitted.
that's not ADHD . People with ADHD would improve - it may take a LOT of time, but it will happen. Quite often they will go to the extreme and come in way too early. My bet would be on Cluster B personality trait e.g. lack of empathy and constant need for attention and validation. ADHD frequently co-occurs with other conditions.
That is not always true and not always with everyone. Many people who have ADHD have unsolvable time blindness. They don't mean to do it but their brain chemistry literally disallows them from not doing so in many cases.
Earlier quoted context omitted.
Throughout my life, what colleagues/friends are unwilling to remark plainly on has been the most telling factor of someone’s character to me.
This can be true I suppose, but equally I have a few friends who practically play characters as if they've resigned themselves to a role in a sitcom. For instance: one of my friends is late to just about everything and treats everyone as if we are on-call. We plainly note this repeatedly, the friend is, I hope, equally frustrated and embarrassed by it, and in spite of this nothing changes. This is obviously a critica…
i.e. If your friends wont remark on your penmanship, who cares? If they wont remark on how you treat service workers at a restaurant, that's probably concerning.
Earlier quoted context omitted.
I think fundamentally, the concern is misplaced. The fact you need to work for wealth is a convention of our constraints. The change in constraints would lead to other means of distribution. It's easy to see if someone who believes more productivity is good would not see making jobs obsolete a real problem. Thew would see us adapting to the new conditions in a relatively short while.
>Thew would see us adapting to the new conditions in a relatively short while. Say ~5 million jobs in the next 10 years are automated away, which industries do those people move to? With college being exorbitantly expensive, that locks out many people from re-skilling in other fields. As people race to other industries, that forces down wages because now there is a larger pool to select from. How do we ensure people…
At the risk of sounding like a longtermist, I think that when all is done, the result will be a net positive one, but it WILL cause strife for many people - probably me included. But I refuse to keep my childern's future hostage because I might have to reskill.
Earlier quoted context omitted.
Human cognition is nothing like AI "cognition." It really bothers me that people think AI is doing the same thing the human mind does. AI is more like a parrot which is trained to give a correct-looking response to any question. The parrot doesn't think, doesn't know what its doing etc, it just does it because it gets a treat every time a "good" answer is prompted. This is why it can't do things like know how many pa…
AI is more like a parrot which is trained to give a correct-looking response to any question. A parrot that writes better code and English prose than I do? I would like to buy your parrot.
Earlier quoted context omitted.
If this is noteworthy, the burden of proof should be on the poster, not the reader to substantiate these claims.
I don't really like this take, as it tries to make being informed somehow not the readers responsibility.
Earlier quoted context omitted.
As a scientist (computational physicist, so plenty of math, but also plenty of code, from Python PoCs to explicit SIMD and GPU code, mostly various subsets of C/C++), I can confirm - Codex is qualitatively better for my usecases than Claude. I keep retesting them (not on benchmarks, I simply use both in parallel for my work and see what happens) after every version update and ever since 5.2 Codex seems further and fu…
I've tried both against similar and haven't found it such a clear cut difference. I still find neither are able to fully implement a complex algorithm I worked on in the past correctly with the same inputs. Not sharing exactly the benchmark I'm using but think about something for improving performance of N^2 operations that are common in physics and you can probably guess the train of thought.
That said, I am puzzled at the algorithms that Claude & GPT "get" and ones that they do not.
(former physicist here. would love to know the kind of things you're working on. email on my profile)
Ronan Farrow here. Andrew Marantz and I spent 18 months on this investigation. Happy to answer questions about the reporting.