Hacker News Reader

Comment by andrewfromx | Hacker News Reader

I can't lie, yes. BUT I used to write the same type summaries with my human brain all the time here too. Is it wrong to get all these points from an AI summary?

I guess you could argue HN might as well just auto summarize every article now!

View on HN

Please don't. HN is a place for humans to write and converse with each other. Bots have never been allowed here and we've banned accounts for posting like this since long before GPT. (I don't want to ban you, obviously)

No its not. You used the tool in useful way, producing result that is helpful to others. Points deserved. Autosumarization would be cool too.

I think we need sort of an etiquette code to disclose when a piece of text is AI-generated. Like when we say "Full Disclosure" or "IANAL"

i can't help think of Data from star trek, longing to be more human like. At some point can the AI be allowed to post with the same equal rights as humans. Picard would fight for Data's rights to comment on hacker news articles.

Presumably he'd be annoyed if you stole Data's work and posted it in your own name though without attribution?

Sorry, I was the person who asked, didn’t mean to get you told off by the attribution police!

I think attributing GPT every time you use it is as stupid as attributing spell check when you use it. It’s a tool, not a person.

Soon it’ll be a click away in the browser and it’ll be used everywhere exactly how you used it.

In fact, pretty soon anyone saying you have to attribute text to GPT will probably be a GPT powered bot! :)

For some reason though, your summary stuck out like a sore thumb. I work a lot with GPT3, I think you get accustomed to how it writes (obviously you can change that with some prompt work).

> Is it wrong to get all these points from an AI summary?

I don't think so, as long as you've vetted it for completeness and accuracy.

That's exactly my point of view. Human-vetted AI content is just content that the human could write themselves, like a fancy autocomplete.

Automatically posting AI content with a bot is a different matter

Yes it is wrong.

You mind explaining how?

If you think it’s useful you’re stealing the work of GPT by posting without attribution, if you don’t you’re wasting our time.

This is where I want to make an omniscient bot that follows you around and gives you a severe electric shock every time you use potentially copyrighting material snippets (I'll be nice and give you a 3 second window to declare fair use).

I will be happy the day copyright is dead and buried under a mountain sized gravestone. As much as you think it protects the small guy, it enabled the rich to buy up portfolios and enslave them to constant payments and holding back the progression of society.

The OP deliberately used another entities work as his own without attribution.

Reply on news.ycombinator.com