Most active commenters
  • petralithic(8)
  • danielbln(3)
  • _Algernon_(3)

←back to thread

I Am An AI Hater

(anthonymoser.github.io)
443 points BallsInIt | 55 comments | | HN request time: 0.823s | source | bottom
1. danielbln ◴[] No.45044286[source]
Yet it is here to stay, won't go away and even if it won't get any better at the useful things it does, it is useful. The externalities are real, some can be removed, some mitigated. If you're a hater and a human, then you don't have to mitigate anything, of course.

Me, I hate the externalities, but I love the thing. I want to use my own AI, hyper optimized and efficient and private. It would mitigate a lot. Maybe some day.

replies(6): >>45044341 #>>45044555 #>>45044591 #>>45044671 #>>45053794 #>>45115936 #
2. danielbln ◴[] No.45044417[source]
AI is information.
replies(3): >>45044507 #>>45044539 #>>45044626 #
3. jerhewet ◴[] No.45044507{3}[source]
AI is "put Elmer's glue on your pizza so the ingredients won't slide off". AI is "three B's in blueberry".

Garbage in, garbage out. Which will always be the case when your AI is scraping stuff off of random pages and commentary on the internet.

replies(1): >>45045156 #
4. zwnow ◴[] No.45044539{3}[source]
Disinformation with more and more propaganda due to being vulnerable to bad actors. There's already evidence on people spreading propaganda through LLMs.
replies(2): >>45044592 #>>45045763 #
5. yifanl ◴[] No.45044555[source]
> Yet it is here to stay, won't go away

Source for this claim? Are you still using Groupon?

replies(5): >>45044611 #>>45044666 #>>45044746 #>>45044805 #>>45044838 #
6. iLoveOncall ◴[] No.45044591[source]
There are plenty of things that are useful and that have gone away. As long as GenAI stays unprofitable, it has every chance of disappearing if it stays as useless as it is right now.
replies(1): >>45044637 #
7. danielbln ◴[] No.45044592{4}[source]
Definitely, AI can be used for terrible things. Doesn't change that it's information and won't go away.
8. sshine ◴[] No.45044611[source]
People are.

Just like crypto.

Just look at the bitcoin hashrate; it’s a steep curve.

9. Mallowram ◴[] No.45044626{3}[source]
Even Shannon knew the limits of information late in career. AI is not information, it's signaling. And it embeds without decipherment or segregating dominance, bias, control, manipulation. The dark matter of language we can't extract.

"Shannon warned in 1956 that information theory “has perhaps been ballooned to an importance beyond its actual accomplishments” and that information theory is “not necessarily relevant to such fields as psychology, economics, and other social sciences.” Shannon concluded: “The subject of information theory has certainly been sold, if not oversold.” [Claude E. Shannon, “The Bandwagon,” IRE Transactions on Information Theory, Vol. 2, No. 1 (March 1956), p. 3.]"

replies(1): >>45053860 #
10. ginko ◴[] No.45044637[source]
There's people running stable diffusion locally on their systems for their own amusement. Do you think that will go away?
replies(1): >>45045695 #
11. brokencode ◴[] No.45044666[source]
You must be living on a different planet if you think the adoption and societal impact Groupon was ever remotely comparable to AI.
replies(1): >>45044719 #
12. myhf ◴[] No.45044671[source]
> the useful things it does

It's weird how AI-lovers are always trying to shoehorn an unsupported "it does useful things" into some kind of criticism sandwich where only the solvable problems can be acknowledged as problems.

Just because some technologies have both upsides and downsides doesn't mean that every technology automatically has upsides. GenAI is good at generating these kinds of hollow statements that mimic the form of substantial arguments, but anyone who actually reads it can see how hollow it is.

If you want to argue that it does useful things, you have to explain at least one of those things.

replies(7): >>45045049 #>>45045236 #>>45045610 #>>45046010 #>>45046790 #>>45047966 #>>45065100 #
13. yifanl ◴[] No.45044719{3}[source]
I assure you as someone working FoH at the time, Groupon's impact on me was far greater than AI ever could be.
replies(1): >>45045385 #
14. dpoloncsak ◴[] No.45044746[source]
The last time we saw a bet from Wall St like we are seeing with AI, was when they bet on the internet.

Do you still use the internet?

replies(2): >>45044873 #>>45045168 #
15. hex4def6 ◴[] No.45044805[source]
Of course it's here to stay. There are models that are --right-now-- great at text-to-speech, speech-to-text, categorization, image recognition, etc etc. Even if progress stopped now, these models would be useful in their current state.

Your argument could just as easily be applied to social networks ("are you still using friendster?") or e-commerce ("are you still using pets.com?). GPT3 or Kimi K2 or Mistral is going to become obsolete at some point, but that's because the succeeding models are going to be fundamentally better. That doesn't mean that they weren't themselves fit for a certain task.

16. sindriava ◴[] No.45044838[source]
Comparing a general technology (AI) to a specific company (Groupon) is a category error. To your point coupons still exist and people use them and Anthropic might not exist in 2 years while AI will.
17. satisfice ◴[] No.45044873{3}[source]
I still use the Internet. The Internet is also a technology that undermines society.
replies(1): >>45045612 #
18. ◴[] No.45045049[source]
19. s1mplicissimus ◴[] No.45045156{4}[source]
If only it were garbage in, garbage out - that would be solvable by better training data. But it's much worse than that, because even if you'd only feed it good stuff, the output would still deteriorate.

pointing index finger at imaginary baloon: pfffffffffft

20. turzmo ◴[] No.45045168{3}[source]
Maybe a better generalization, the last time [bubble] happened, do you still use [bubble]?

Depends on the nature of the bubble, doesn't it?

replies(1): >>45045529 #
21. schwartzworld ◴[] No.45045236[source]
It's bad at

- Actually knowing things / being correct - Creating anything original

It's good at

- Producing convincing output fast and cheap

There are lots of applications where correctness and originality matter less than "can I get convincing output fast and cheap". Other commenters have mentioned being able to vibe-code up a simple app, for example. I know an older man who is not great at writing in English (but otherwise very intelligent) who uses it for correspondence.

replies(1): >>45047431 #
22. brokencode ◴[] No.45045385{4}[source]
I’m talking about how it impacts society in general, not you specifically. Also, I don’t think you appreciate how deeply AI will affect your life in the future.

The next time you get a CT for example, it might be an AI system that finds a lung nodule and saves your life.

Or for a negative possibility, consider how deepfakes could seriously degrade politics and the media landscape.

There are massive potential upsides and downsides to AI that will almost certainly impact you more than a coupon company.

23. TeMPOraL ◴[] No.45045529{4}[source]
Indeed. Therefore, the past bubbles most similar to AI are the one around Internet, and the earlier one around electricity.
replies(1): >>45050614 #
24. petralithic ◴[] No.45045610[source]
> doesn't mean that every technology automatically has upsides

Who said "every technology?" We're talking about a specific one here with specific up and downsides delineated.

25. petralithic ◴[] No.45045612{4}[source]
You are free not to use it if that is what you believe.
replies(2): >>45045820 #>>45045937 #
26. petralithic ◴[] No.45045630[source]
I don't understand this sentence. What is "operating the arbitrary?"
replies(1): >>45051171 #
27. brian-armstrong ◴[] No.45045695{3}[source]
100% yes, at least down to a rounding error. If people stop pushing billions into training new versions, then the novelty will wear off very quickly. There are still many constraints on what it can do and people will generally lose motivation when they start finding those invisible boundaries on its capabilities. It'll be effectively a dead pursuit.
28. justsomejew ◴[] No.45045763{4}[source]
I think you are a real person, still you sound like a broken record.. "disinformation..", "propaganda".. "bad actors"..

You are the "bad actors", pumpkin. Worse than the other ones.

replies(1): >>45049488 #
29. Timwi ◴[] No.45045937{5}[source]
You are also free to give away all your money and not participate in capitalism if you wish.

Wait, are you sure?

replies(2): >>45046453 #>>45049424 #
30. Semiapies ◴[] No.45046010[source]
And they're always desperately insisting that won't go away and you can't escape it. It stinks a lot of Big Lie techniques.
replies(1): >>45053824 #
31. petralithic ◴[] No.45046453{6}[source]
Yes, if you don't want to participate in capitalism, you're free to live in the woods and homestead.
replies(3): >>45047058 #>>45048862 #>>45054198 #
32. int_19h ◴[] No.45046790[source]
It is extremely good at translating things from one language to another. Even if nothing else comes out of the AI bubble, this thing alone is a massive upside.
replies(1): >>45047961 #
33. satisfice ◴[] No.45047058{7}[source]
No, you are not free to do that. There are billions of humans on the planet. We all can't "live in the forest" without immediately destroying the forests.

I wish you'd try thinking for at least five seconds before commenting. If you are here, then you must be smart-- so, use your brain, man.

replies(2): >>45049726 #>>45050746 #
34. const_cast ◴[] No.45047431{3}[source]
Those applications are actually abysmally rare - it's just that we've created a society where businesses are just... allowed to externalize all costs.

Being wrong or lying is almost universally bad and unproductive. But making money has nothing to do with being productive - you can actively make the world worse and make money. Ask RJ Reynolds.

replies(1): >>45052158 #
35. SpaceNoodled ◴[] No.45047961{3}[source]
Sure, LLMs seem to be really good at outputting things that aren't trivially verifiable by the user.
36. joquarky ◴[] No.45047966[source]
Brainstorming.
37. _Algernon_ ◴[] No.45048862{7}[source]
https://iea.org.uk/wp-content/uploads/2021/12/mister-gotcha-...
replies(1): >>45050753 #
38. Gud ◴[] No.45049424{6}[source]
Not really an option, is it? Point me to the forest where this is allowed
39. zwnow ◴[] No.45049488{5}[source]
So we are just ignoring this issue?
40. jacquesm ◴[] No.45049726{8}[source]
> If you are here, then you must be smart

Citation needed...

41. Jensson ◴[] No.45050614{5}[source]
> the past bubbles most similar to AI are the one around Internet

I'd say crypto is more similar, the internet and telephones and rails and roads and electricity is around connective infrastructure, AI and crypto is around compute. Connective infrastructure is almost always useful, local computed things is harder to motivate hype for because its usefulness isn't as apparent as adding more connectivity.

replies(1): >>45050902 #
42. petralithic ◴[] No.45050746{8}[source]
Who said we all? If you specifically don't want to live in a capitalist society, you don't have to, there are lots of homesteaders even in the US who live off the land, you can be like them. I'm not being sarcastic, it's an actual suggestion.
43. petralithic ◴[] No.45050753{8}[source]
I see this meme often but it's true, there's lots of options, not just one society that you cannot leave. Lots of countries on earth that one can move to, as people immigrate already.
replies(1): >>45051390 #
44. dpoloncsak ◴[] No.45050902{6}[source]
Crypto STILL hasn’t seen widespread user adoption like ChatGPT has.

My CIO never asked about blockchain technology, but he sure as hell is asking about AI

45. Mallowram ◴[] No.45051171{3}[source]
words, symbols, sentences, tokens, all are arbitrary. They stay arbitrary unless there is an irrefutable target like a tumor. This is the basis of CS
46. _Algernon_ ◴[] No.45051390{9}[source]
Talk about missing the point…
replies(1): >>45052470 #
47. schwartzworld ◴[] No.45052158{4}[source]
> Being wrong or lying is almost universally bad and unproductive

Sure, but there are cases where rightness isn’t a thing.

Don’t get me wrong I’m not an AI Stan, it has real problems, but it’s also not going anywhere. Eventually the bubble will pop and we’ll see which applications of AI turned out to be useful and which didn’t.

48. petralithic ◴[] No.45052470{10}[source]
More like just presenting a meme isn't a real argument, so there's not really any point you're making to miss.
replies(1): >>45054148 #
49. ◴[] No.45053794[source]
50. lif ◴[] No.45053824{3}[source]
This ship is unsinkable!

-Aitanic

51. lif ◴[] No.45053860{4}[source]
Signaling? That is correct. Also, am willing to place a _very_ long bet that clay tablets will be around longer.
52. _Algernon_ ◴[] No.45054148{11}[source]
If you practiced some media literacy, you'd realize that (a) there is a point to that meme and (b) that there is a reason why you are on the receiving end of it.
replies(1): >>45055402 #
53. petralithic ◴[] No.45055402{12}[source]
Just saying it doesn't make it so. I can post similar memes and say "haha that's you" while pointing at you and citing your lack of media literacy when you disagree, that doesn't mean it's a cogent argument. Maybe there is an argument in that meme, but you're not specifying it and therefore I will not take it as a serious comment.
54. gitaarik ◴[] No.45065100[source]
It's just a summarizing search engine. Instead of you manually looking up multiple pages and reading them and coming to some conclusion that allows you to move to the next step in the process, you can let the AI do that manual work for you.

But you shouldn't expect it do take over your actual thinking, because it doesn't actually think. So it's just another tool in the toolbox that can be useful for some applications, but not for all. If you use it for the appropriate tasks, it can be very helpful. If you try to do everything with it, you'll be disappointed.

55. lproven ◴[] No.45115936[source]
> Yet it is here to stay, won't go away and even if it won't get any better at the useful things it does, it is useful.

"Its power seems inescapable. So did the divine right of kings." — Ursula K. Le Guin