Thread

1. lucfranken · 1791024804 · focus · HN ↗

   Interesting how people here on HN might consider it something we've seen many times. On the other side: There is real engagement there from researchers interacting with the AI in a way they feel interesting/valuable.

   We should not forget the way broader levels of knowledge on topics around the world compared to our own small world.

   1. MattGaiser · 1791024929 · focus · HN ↗

      Yeah, I do not get the dismissiveness given that the researchers themselves were not as dismissive.

      > The connection was a “little bit far-fetched,” she says. “But I could imagine a first-year Ph.D. student coming up with exactly those same questions and ideas.”

      This one in particular made it seem like a very ordinary interaction.

      1. itsalwaysgood · 1791026869 · focus · HN ↗

         It's kind of misleading information. Most who have worked with AI can see issues with what the AI Engineer created, and how what he's doing can't be 'scaled up' to society.

         Everyone in the article seems desperate for attention and engagement, and it makes some readers uncomfortable: because of the awkwardness, and because this guy is spreading a 'bad idea' that probably shouldn't be replicated.

         Imagine if everyone starts spamming emails? The researchers are amused this time, but I guarantee they wouldn't be if their inbox filled up with agent emails.

         Not to mention the way the agents reached out for help is probably a bad idea: asking someone with 30 years of experience to apply all their knowledge to your small, specific problem in another knowledge domain, with no other detail or context puts all the mental burden on the unsolicited email recipient.

         1. MattGaiser · 1791027294 · focus · HN ↗

            There is already mass spamming emails. I already get hundreds of emails per day that are spam as I post my emails publicly for people to get in touch. If you want to truly and maliciously spam, you can do that.

            The fix is already AI triage.

            That is the current state of email today.

            So why couldn’t this scale unless for some reason our reaction to this is to shut off email and not build better classifiers?

            1. itsalwaysgood · 1791027850 · focus · HN ↗

               Personally, I don't care what the guy does. It's entertainment for him, and obviously the people replying to his emails were entertained. I doubt they would continue to be entertained if more agents emailed them. I'm just pointing out issues with the article, why people might be upset.

               Also, Hackernews tries to hold itself to a certain standard. Reading an article about an AI Engineer who gives an AI agent access to a tool, with clear directive and access to use the tool, who is then later confused when the agent uses the tool in a slightly different way, suggests the article is below Hackernews standards.

               To answer your question though: yes you absolutely could fight spam with AI on your side.

               Should we go down that path as a society, just for amusement? Essentially saying: yes, someone should pay for AI service to combat AI spam, just so we can continue to use email.

Unofficial Hacker News client; not affiliated with Y Combinator.