I don't trust anyone who claims that LLMs today are superhumanly intelligent. All they do is perform compute-intensive brute-force attacks on the problem/solution space and call it 'reasoning', all while subsidising the real costs to capture the market. So much SciFi BS and extrapolation about a technology that is useful if adopted with care. This technology needs to become a commodity to destroy this aggregation of…
Sam Altman may control our future – can he be trusted?
271–280 of 1001 posts
Re: Sam Altman may control our future – can he be trusted?
#272Ronan Farrow here. Andrew Marantz and I spent 18 months on this investigation. Happy to answer questions about the reporting.
Re: Sam Altman may control our future – can he be trusted?
#273Ronan Farrow here. Andrew Marantz and I spent 18 months on this investigation. Happy to answer questions about the reporting.
As a reader, am I supposed to infer anything about evidentiary weight from these stylistic choices? When a single anonymous source's testimony is presented in a "declarative" narrative style like here (with the attribution in a less prominent position), should we read that as reflecting high confidence on your end (perhaps from additional corroboration not fully spelled out)? And does the fact that Altman’s non-recollection appears in parentheses carry any epistemic signal (e.g. that you assign it less evidentiary weight)? Or is that mostly a matter of (say) prose rhythm?
Re: Sam Altman may control our future – can he be trusted?
#274I don't trust anyone who claims that LLMs today are superhumanly intelligent. All they do is perform compute-intensive brute-force attacks on the problem/solution space and call it 'reasoning', all while subsidising the real costs to capture the market. So much SciFi BS and extrapolation about a technology that is useful if adopted with care. This technology needs to become a commodity to destroy this aggregation of…
Your brain is performing "compute-intensive brute-force attacks on the problem/solution space" as you read this very sentence. You trained patterns on English syntax, structure, and semantics since you were a child and it is supporting you now with inference (or interpretation). And, for compute efficiency, you probably have evolution to thank.
Re: Sam Altman may control our future – can he be trusted?
#275Ronan Farrow here. Andrew Marantz and I spent 18 months on this investigation. Happy to answer questions about the reporting.
The statements around the sexual abuse allegations seemed to be the most puzzling to me - his sister’s allegations and claims of underage partners because he has a tendency to hook up with younger partners. It does seem like this piece gives him a pretty clean bill of health in that matter - I guess would you be able to talk about how you investigated? Did you do any extra investigations into Annie’s allegations? It…
Re: Sam Altman may control our future – can he be trusted?
#276Earlier quoted context omitted.
[flagged]
> Struck a nerve? No need to be petty. They have a point. We did this with the words racist and fascist. Overinclusion diluted the term and gave cover for the actual baddies to come in. I'm not sure debating who is and isn't a sociopath is as useful as, say, the degree to which Sam is a liar (versus visible).
Re: Sam Altman may control our future – can he be trusted?
#277Re: Sam Altman may control our future – can he be trusted?
#278I don't trust anyone who claims that LLMs today are superhumanly intelligent. All they do is perform compute-intensive brute-force attacks on the problem/solution space and call it 'reasoning', all while subsidising the real costs to capture the market. So much SciFi BS and extrapolation about a technology that is useful if adopted with care. This technology needs to become a commodity to destroy this aggregation of…
Your brain is performing "compute-intensive brute-force attacks on the problem/solution space" as you read this very sentence. You trained patterns on English syntax, structure, and semantics since you were a child and it is supporting you now with inference (or interpretation). And, for compute efficiency, you probably have evolution to thank.
Re: Sam Altman may control our future – can he be trusted?
#279Earlier quoted context omitted.
There's a time and a place for everything, and rejecting popular media as "lowest common denominator" is the most uninspired form of cultural elitism. Is it cynical to want your to make a profit? Or for it to make enough profit to subsidize other projects? Is it cynical to make something accessible so more people who watch it are able to enjoy it? I agree that it's embarrassing and feels crass when movies both try to…
> There's a time and a place for everything Yeah, I'm saying professional communication isn't the place for Marvel references, and that those who choose to include references to those movies in their professional communications are revealing something about their media tastes. If I'm at a Michelin star restaurant I don't want to be served a ballpark hotdog.
Re: Sam Altman may control our future – can he be trusted?
#280For those curious about how sama got to where he got and stayed on top for so long, I recommend you read the book: The Sociopath Next Door by Martha Stout. I am fairly confident when I say this -- sama is a sociopath. I don't know how anyone with solid intuition could even come to any other conclusion than the guy is deeply weird and off-putting. Some concepts from the book: > Core trait: The defining characteristic…
> OpenAI is too important to trust sama with.
...wat? They made a chat bot. How can that possibly be so existentially important? The concept of "importance" (and its cousin "danger") has no place in the realistic assessment of what OpenAI has accomplished. They haven't built anything dangerous, there is no "AI safety" problem, and nothing they've done so far is truly "important". They have built a chat bot which can do some neat tricks. Remains to be seen whether they'll improve it enough to stay solvent.
The whole "super serious what-ifs" game is just marketing.