> Ever wondered why 80% of vehicles and corporate pub looks the damn same, and how everything is sanitised and monocultured? Every new building is minimalist in design, and that's not a result of artistic choice. Everything on the corpo Web is corporate marketing, and says nothing of value. One thing 'AI' cannot do is produce anything original or inspiring. It will generate something derivative of this blandness.
While it is true that AI is training on bland and thereby producing more bland, much of the bland has an older history — standardization curbed creativity long before AI did. Websites shifted to corporate bland when everybody started using templates and drag-and-drop editors in the 2010s. Cars started to look the same when vehicle manufacturers shifted to a standard chassis and body that already carried crash-test approval rather than designing new ones to be approved. AI is but the latest instance of this trend but potentially far more damaging because it can spread the bland further and much more quickly.
> The Tech Bros are psychopaths. More specifically, if you took their own words at face value, the Tech Bros are Epstein-class psychopaths, openly committed to an anti-human ideology, who boast, on stage, about how they spend obscene amounts of money on something they believe might destroy humanity.
Trying to codify my thoughts on LLMs to someone less technical conjured similar thoughts for me. If anybody but Amodei/Altman made the claims they're making ("we're worth 2 trillion dollars because we might end the world") they'd be written off as an utter crackpot.
Well, hyperbole. But you're right that I didn't go about commenting in the most constructive way.
To be a little more generous, my understanding is that many an e/acc would roughly agree:
- AI may well put most people out of a job
- AI may well become, for lack of a better word, Skynet (even if by accident)
- It is fair to aggressively scrape the internet, pirate terabytes of copyrighted works (hosted illegally), and destructively scan books, even if rare, to further AI research
- Software and its documentation need not be written nor reviewed by humans
And I have a feeling the average Joe, who treats "ChatGTP" as a way to generate a cocktail recipe or write an funny poem on occasion, would see these as utterly deranged and totally inexcusable.
I'm not even arguing whether it is or not - I just suspect that's how most people would see it.
> Ever wondered why 80% of vehicles and corporate pub looks the damn same, and how everything is sanitised and monocultured? Every new building is minimalist in design, and that's not a result of artistic choice. Everything on the corpo Web is corporate marketing, and says nothing of value. One thing 'AI' cannot do is produce anything original or inspiring. It will generate something derivative of this blandness.
While it is true that AI is training on bland and thereby producing more bland, much of the bland has an older history — standardization curbed creativity long before AI did. Websites shifted to corporate bland when everybody started using templates and drag-and-drop editors in the 2010s. Cars started to look the same when vehicle manufacturers shifted to a standard chassis and body that already carried crash-test approval rather than designing new ones to be approved. AI is but the latest instance of this trend but potentially far more damaging because it can spread the bland further and much more quickly.
> The Tech Bros are psychopaths. More specifically, if you took their own words at face value, the Tech Bros are Epstein-class psychopaths, openly committed to an anti-human ideology, who boast, on stage, about how they spend obscene amounts of money on something they believe might destroy humanity.
Trying to codify my thoughts on LLMs to someone less technical conjured similar thoughts for me. If anybody but Amodei/Altman made the claims they're making ("we're worth 2 trillion dollars because we might end the world") they'd be written off as an utter crackpot.
Well, hyperbole. But you're right that I didn't go about commenting in the most constructive way.
To be a little more generous, my understanding is that many an e/acc would roughly agree:
- AI may well put most people out of a job
- AI may well become, for lack of a better word, Skynet (even if by accident)
- It is fair to aggressively scrape the internet, pirate terabytes of copyrighted works (hosted illegally), and destructively scan books, even if rare, to further AI research
- Software and its documentation need not be written nor reviewed by humans
And I have a feeling the average Joe, who treats "ChatGTP" as a way to generate a cocktail recipe or write an funny poem on occasion, would see these as utterly deranged and totally inexcusable.
I'm not even arguing whether it is or not - I just suspect that's how most people would see it.
I agree with the article.
When you see extremes like this and the AI grifter bros' hype, you know the truth is somewhere in the middle.
It's Stallman vs Corporate, and Torvalds is the solution point, in-between.