"The pattern that works is using agents to produce deterministic artifacts, then forcing execution through those constraints."
-
People rarely need to comment, but we do anyway.
It's not surprising that a 1Password customer -- who read the toot and then was blocked from reading the details -- would react negatively.
@RandomOxen @1password people are welcome to jump to conclusions. Doesn’t make them right
️ -
"The pattern that works is using agents to produce deterministic artifacts, then forcing execution through those constraints." Tido Carriero, VP of Engineering at Cursor.
At 1Password, we applied agentic tooling to B5, our multi-million-line Go monolith, to help plan and execute a production refactor. Here's what we learned: https://1password.com/blog/what-we-learned-using-ai-agents-to-refactor-a-monolith
@1password I cancelled my 1Password sub when you guys jumped on the crypto bandwagon a few years ago. Not doing a very good job at convincing me to come back.
-
"The pattern that works is using agents to produce deterministic artifacts, then forcing execution through those constraints." Tido Carriero, VP of Engineering at Cursor.
At 1Password, we applied agentic tooling to B5, our multi-million-line Go monolith, to help plan and execute a production refactor. Here's what we learned: https://1password.com/blog/what-we-learned-using-ai-agents-to-refactor-a-monolith
@1password Ah. I see. Well you are definitely not a consideration for moving password managers. Go fuck yourselves.
-
@1password Bye. Dropping this service immediately. Good job giving me a reason to move my family plan away. Time to check on ChiPass progress.
@cararemixed @1password slowly but steadily we're progressing towards the first release
-
@RandomOxen @1password people are welcome to jump to conclusions. Doesn’t make them right
️I think you might be thinking naívely here. Or I might be. It's hard to say.
When information is being withheld, people -- rightly -- use their knowledge and experience to fill in the gaps.
They shouldn't be expected to be silent, even if they fill those gaps incorrectly.
People don't operate like that.
-
They seem to require cookies in order to show their blog post, so it's not surprising that some folks didn't read the blog.
You have the option to reject cookies and still read the blog as soon as you open the page.
-
I think you might be thinking naívely here. Or I might be. It's hard to say.
When information is being withheld, people -- rightly -- use their knowledge and experience to fill in the gaps.
They shouldn't be expected to be silent, even if they fill those gaps incorrectly.
People don't operate like that.
@RandomOxen @1password I’m not seeing a cookie block. And even if it’s blocked like that, everyone here seems to be claiming to be a customer, so I thought they might owe the company a little more grace than just reading the headline. But again, you’re right, people can and will jump to conclusions. That still doesn’t make them right.
-
@cararemixed @1password slowly but steadily we're progressing towards the first release
@whitequark @1password I've got a local clone already and have been following. This just triggered an earlier migration than I planned.
-
@whitequark @1password I've got a local clone already and have been following. This just triggered an earlier migration than I planned.
@cararemixed @1password probably the biggest challenge right now is delivering high-assurance release builds for multiple platforms. it's happening but it is taking quite a bit of effort to build that infrastructure (and associated things like signing with an HSM) on top of Codeberg
-
@RandomOxen @1password I’m not seeing a cookie block. And even if it’s blocked like that, everyone here seems to be claiming to be a customer, so I thought they might owe the company a little more grace than just reading the headline. But again, you’re right, people can and will jump to conclusions. That still doesn’t make them right.
I suspect now that it's my block on site data / local storage that prevents the page from loading.
-
"The pattern that works is using agents to produce deterministic artifacts, then forcing execution through those constraints." Tido Carriero, VP of Engineering at Cursor.
At 1Password, we applied agentic tooling to B5, our multi-million-line Go monolith, to help plan and execute a production refactor. Here's what we learned: https://1password.com/blog/what-we-learned-using-ai-agents-to-refactor-a-monolith
@1password you’re on very shaky ground. This wouldn’t make me cancel yet, but it will make me start evaluating some pros and cons of other password managers as a “just in case”. I’ll want to know where to go when I hear of 1Password’s first major breach.
-
"The pattern that works is using agents to produce deterministic artifacts, then forcing execution through those constraints." Tido Carriero, VP of Engineering at Cursor.
At 1Password, we applied agentic tooling to B5, our multi-million-line Go monolith, to help plan and execute a production refactor. Here's what we learned: https://1password.com/blog/what-we-learned-using-ai-agents-to-refactor-a-monolith
@1password WTH
-
@1password I guess I need to find a replacement password manager asap. Well done.
@glebd @1password Same here. I’m looking for something cross-platform for Linux, MacOS, iOS and Windows, probably self-hosting on my NAS via Tailscale. What’s annoying is that I only recently moved my wife and stepdaughter onto 1Password.
-
@glebd @1password Same here. I’m looking for something cross-platform for Linux, MacOS, iOS and Windows, probably self-hosting on my NAS via Tailscale. What’s annoying is that I only recently moved my wife and stepdaughter onto 1Password.
@RichardBuckle I’m looking at BitWarden or Proton Pass. Leaning towards BitWarden’s EU instance, bitwarden.eu (it can be self-hosted but I can’t be bothered tbh).
-
@NE555 @1password @protonprivacy I'm sure the proton devs use LLMs as well for development, so I would question if that makes a difference
@klausi @1password @protonprivacy Really? Would be nice if @protonprivacy could transparently communicate that.
I just looked into the repos on GH and to me it doesn't seem that there are any hints that they are using LLMs. A web search was tough (as Lumo and Scribe dominate the results), but also didn't reveal anything.
If #Proton really use LLMs to write their code, then why would we pay for that? I am paying for well crafted services, not LLM BS. I hope your assumption is not correct.
-
R relay@relay.mycrowd.ca shared this topic
-
@1password there are a whole lot of responses to this that show me they didn’t read the article. Thanks for sharing your experiences!
Sometimes people can possess the same information as you and still have a different opinion.
How mad is that?
-
@1password So, the 1password backend now operates on AI generated code which is known to introduce security vulnerabilities.
You are aiming for a LastPass type of incident?
Seems like I have to evaluate a new password manager…@virtualmarc @1password
Luckily, #vaultwarden seems to be unmolested so far.
