logoalt Hacker News

Dots: Always-on agents

446 points • by alvis • yesterday at 5:07 PM • 337 comments • view on HN

Comments

esafak • yesterday at 6:34 PM

I have not tried it yet but this looks as risky as openclaw, which I also won't use. What if it does something I would not have approved and I only found out about it later? Knowing how often agents go off the rails when I'm coding, I would hesitate to let one do other tasks. I would prefer to white list tasks one at a time as I gained trust.

petesergeant • yesterday at 6:10 PM

I spend literally all my work day, and a good bit of my personal time, talking to agents, getting them to do things on my behalf. Almost always pretty tightly sandboxed. I just don't understand how people using these things haven't had catastrophic failures yet.

I minted what I thought was a minimal-permission Github token for a single action, and the agent I gave it to discovered it had more permissions than I thought, and made use of those permissions. Who is trusting these things with write access to their lives?

➕ show 2 replies
simianwords • yesterday at 5:57 PM

I hear that Dots don't count towards usage, is that true?

gnarlouse • yesterday at 6:26 PM

I love stuffing eldritch horror inside a teddie bear plushie. So perfectly a caricature of modern tech-dystopianism.

God I want the fucking market to crash

glub • yesterday at 5:19 PM

So essentially, another attempt at giving codex to regular people.

etchalon • yesterday at 5:34 PM

At some point a designer is going to come up with a visualization for an agent that isn't "amorphous character blob".

everfrustrated • yesterday at 6:17 PM

Lolz https://dot.com goes to xAI's GrokBot (the competition)

jdw64 • yesterday at 5:19 PM

I heard they use GPT Space (like Notion) together with Slack and use it like a human colleague, but watching the actual demo video, the speed is so slow it's shocking..

stronglikedan • yesterday at 6:55 PM

Holy crap that is some expensive food in the demo video. These people are truly disconnected from reality.

AlfredBarnes • yesterday at 5:46 PM

Muse and now Dot's emergence in popularity makes me think Apple's "lil Finder" will be their take on it.

VonLuderitz • yesterday at 7:22 PM

The 3D print in 0:06 is failing LOL

deno • yesterday at 7:24 PM

Most predictable announcement ever. I would prefer if they just removed scheduled tasks limits instead, at least in Work mode. Just use my usage for crying out loud.

i_love_retros • yesterday at 6:52 PM

Meta's weird keychain ai thing and now this, both going for that cute vibe so we forget how scummy these companies and their owners are.

einpoklum • yesterday at 6:14 PM

I'll take always-off agents please.

_nakul • yesterday at 10:44 PM

[flagged]

Conol_ai • yesterday at 6:40 PM

[dead]

ranyume • yesterday at 5:23 PM

It's great to see that my personal assistant can stop working and maybe report me to the police if the company disagrees with what I'm doing!

➕ show 2 replies
mhog_hn • yesterday at 5:22 PM

[flagged]