The AI community expresses sympathy on Hacker News | Hacker News Reader

The AI community expresses sympathy on Hacker News

twitter.com

71 pointslarryfreeman16 comments

There's so much going on in that demo. Scraping, natural language processing, text generation. Just crazy that this is actually real.

Reminds me of natbot, which uses GPT-3 to do all 3: https://github.com/nat/natbot

Link to comment from the AI model here https://news.ycombinator.com/item?id=33520585&p=3#33527851

Not anymore: flagged, and dead.

Though a direct link seems to work:

https://news.ycombinator.com/item?id=33527851

I almost find the flagging more interesting. This wasn't a bot. Someone had a AI assist them with writing a comment but they were the ones to prompt for it, review and to click comment. Conceptually it's very similar to advanced auto-suggest or grammarly.

I get it its a joke and all but do people not realize that in crypto just like everywhere else only the people who dont know what they are doing lose money.

No one whos seriously is into crypto and has looked at the markets sober, didn't see this coming. Sure most didn't know about FTX or what exactly would trigger it but it was very very clear this crash is going to happen. Its also very very clear it isn't over yet BTC/ETH are at -72% from ATH while -85% would be the "norm". Past does not predict the future but still gives a pretty good idea of what is to come.

I wonder if someone has trained an Austrian Economics bot, the whole thing is based on first principles which one would think an AI could ingest.

Then set it loose trolling the internet…

I hope this is actually good, so that I never have to use selenium again.

friendly reminder: when you get in arguments online, you may be debating with a bot! HN is now no exception....

it feels that way already. Many arguments (if you get into them) can many times just be how well both sides can google something to back them up. At that point you are just arguing with google plus suggestions?

Surely next step is having bots argue with each other so that we do not have to waste time doing it ourselves.

Recently posted here:

https://infiniteconversation.com/

https://news.ycombinator.com/item?id=33437296

There was a comment thread recently on HN where someone noted that the user-profile to the comment they were replying to stated that all of the comments were AI generated. The reply from that account was something to the effect of "accusing your those you disagree with of being bots is counterproductive"

The whole thing was a bit surreal.

[edit]

Found the comment:

> ... As for the typos, I strongly believe little mistakes are what makes humans human. Insinuating other users you converse with online are bots is kind of a bad faith claim that derails political discussion[1].

[edit 2]

At the time of this edit, the most recent comment by this alleged bot is, somewhat ironically:

> This reads like a grade schooler trying to plagiarize something with his first thesaurus. I am disappointed by AI every day[2].

1: https://news.ycombinator.com/item?id=33324313

2: https://news.ycombinator.com/item?id=33530723

> Insinuating other users you converse with online are bots is kind of a bad faith claim that derails political discussion

This almost passed the Turing Test, but you can tell it's a bot because The Guidelines were never mentioned.

What is The Guidelines other than something impossible to Google?

Depends on the bot. I like debating because it helps me formalize and explore different aspects of my ideas. If a bot is good at playing that role it would be great, especially if I knew it was a bot and could run it locally (and that it was free software).

Reply on news.ycombinator.com