Responsibility is only mentioned in the context of who to contact regarding this. It’s basically do not bother my wife I cheated on, she only found out about it from the news like all of you.
Responsibility implies consequences and as far as I can tell there are none so far, other than minor embarrassment.
This is an interesting request. Every claim of LLM being wrong has to be proven.
How about the opposite? Share a normal session (by normal, I mean something longer than, say, 4 turns, not absolutely trivial) where LLM doesn’t do a single silly thing. Including in thinking portions.
> This is an interesting request. Every claim of LLM being wrong has to be proven.
Providing even a minimal set of details is important because “an LLM” could mean anything from the Google search header to Fable or Astra.
It’s like someone making a claim that they tried a diet and didn’t lose weight, therefore all diets are bad and shouldn’t be trusted. The details of the diet, how long they tried it, how strictly they followed it, and the rest of their situation are pertinent.
There is an audience who demands content like that: People who want do hate diets and believe they don’t work will eat that commentary up because it levels up the belief they want to hold. Everyone with critical thinking skills or experience with dieting knows that one person claiming diet failure means nothing.
We know diets are a set of tools and they’re not perfect, but they are used successfully by people who can understand their limitations. Now replace diet with LLM and my statement is the same
> How about the opposite? Share a normal session (by normal, I mean something longer than, say, 4 turns, not absolutely trivial) where LLM doesn’t do a single silly thing. Including in thinking portions.
How about Terence Tao sharing his ChatGPT logs having ChatGPT work with him through understanding the Jacobson conjecture?
Absence of such a claim is a commentary on its own, intended or not. Economic activity is almost always presented as a positive thing. It’s said to create jobs, generate wealth, increase production, etc. And here most of the things on the list moved massive amounts of capital without producing any of the positive effects. We get an odd WWW proposal that did actually produce those effects, but for each of those we also get a dozen of blockchain links. I’d say same goes for AI. While it has more utility than blockchain ever did it’s hard to say it was net positive overall to date.
How much money? I mean, the natural answer is full price but that’s not the amount the author gets from a regular sale. There’s a whole range where it’s cheaper to freebooters and still more profitable to the author.
I’ve seen an argument that buying an album from a musician (or a physical media or merch) and pirating the rest of their library in most cases more profitable to the musician than streaming.
I suppose there’s a similar dynamic in publishing where the overhead created by all the intermediaries is significant compared to the actual work done by the author, editor, designers, and so on.
If you send the author half of what the book costs, there's an almost 100% chance this is more than they would have made if you had bought the book. If you send the author 20%, that's roughly what most authors earn per book sale. It can be substantially higher or lower depending on a bunch of factors, but I think it's a reasonable rule of thumb.
> I’ve seen an argument that buying an album from a musician (or a physical media or merch) and pirating the rest of their library in most cases more profitable to the musician than streaming.
Streaming is a particularly brutal example for musicians. Book authors make quite a bit more per book sale than musicians make per stream.
Wouldn't it be cool if we could get to the point where authors (and artists generally) are making more from voluntary direct payments than they are from the platforms?
It would be a great way to demonstrate how broken the funding model is, and all of that money would be money that didn't contribute to the surveillance/censorship tech that copyright is used to justify.
By this measure tweeting “end your life” at no one in particular should be classified as a weapons of mass destruction.
There’s intent to cause harm. If people actually do, it would substitute achievement of the intent. The mechanism is novel, unlike knives and bullets. Maybe hit rate is a bit low but still, the potential number of targets makes it almost a certainty it would work.
—
We learned back in 80s—even earlier—that mixing data and executable code is not a good idea. It took some decades to move onto a different approach. Now we’re back to it with LLMs. It’s not a novel problem. The results are very much predictable.
Nobody thinks a single tweet telling someone to commit suicide has a significant probability of success. It manifestly does not or else we would be an extinct species by this point.
People have been convicted for using words to convince other people to commit suicide.
Broad appeal? You can’t do anything good aiming for broad appeal. People like too different things, they all don’t mash together well. If you add to many tings that appeal to different people you end up with incoherent mess. Instead of appealing to everyone you end up pissing off everyone in many different ways.
Look at the Ring of Power. Did it have broad appeal? Maybe, first 2 episodes brought in 25 million viewers. The whole first season was watched by 150 millions, despite season completion being ~40%. For reference, X-Files pulled in more only in 1 of its 11 seasons with most under 20 mil, and Friends lowered in 20-25 mil with only one season being about 30 mil. So the numbers are impressive but do you know anyone who loves the Rings of Power? Anywhere near to how people love SG-1?
Responsibility implies consequences and as far as I can tell there are none so far, other than minor embarrassment.
reply