"The pattern that works is using agents to produce deterministic artifacts, then forcing execution through those constraints."
-
@1password many posters apparently think otherwise but AI also catches many bugs.
@karl @1password Alas, AI hyping and snake oil salesmanship by tech bros, as well as general misuse of AI in software development, have left a very bad taste in everyone’s mouth.
-
Actually, a good summary of the lessons. From the business and engineering perspectives, I have a few questions; How do you measure ROI? When is it advantageous for engineers to leverage LLMs, when would it be more beneficial to hire a new FTE?
Finally, how do you maintain engineer motivation, especially when LLMs can handle a significant portion of their work? And how do you ensure a consistent influx of junior engineers while also fostering their continued learning?
At the end of the day, LLMs are trained by data created by engineers. No engineers left == no data for LLMs to train.
Compared to others in the comments, I'm actually happy to see how you think about using LLMs within the organization.
No. LLMs cannot do those jobs, those folks are paid to make software that works right
-
@1password many posters apparently think otherwise but AI also catches many bugs.
-
No. LLMs cannot do those jobs, those folks are paid to make software that works right
If you're still clinging to the idea that LLMs are a bad tool for engineers, you're going to get left behind.
️ -
@1password So where do I switch to that does not use LLMs for this? So sad that so much once great software gets worse these days.
@yatil @1password There's chipass.
ChiPass
Codeberg is a non-profit community-led organization that aims to help free and open source projects prosper by giving them a safe and friendly home.
Codeberg.org (codeberg.org)
"KeePassXC asks us to be skeptical of them if we are skeptical of LLMs. This is a convincing argument."
-
"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 Fucking hell. You are using LLM slop code now? Great, now I need to migrate.
-
If you're still clinging to the idea that LLMs are a bad tool for engineers, you're going to get left behind.
️To elaborate a little more; automation always helped people to write better and more code. LLMs are just the next generation of automation tools. Besides writing code, in my experience it can help engineers understand and learn about certain topics.
I believe engineers who effectively integrate LLMs into their workflow will be more productive and produce better code compared to those who don't. This is similar to how using Ansible for automation helps with faster and more reliable builds, or how auto completion in an IDE improves code quality.
Furthermore, the entry barrier into IT, engineering, and coding fields has considerably lowered, which I consider a positive development.
-
To elaborate a little more; automation always helped people to write better and more code. LLMs are just the next generation of automation tools. Besides writing code, in my experience it can help engineers understand and learn about certain topics.
I believe engineers who effectively integrate LLMs into their workflow will be more productive and produce better code compared to those who don't. This is similar to how using Ansible for automation helps with faster and more reliable builds, or how auto completion in an IDE improves code quality.
Furthermore, the entry barrier into IT, engineering, and coding fields has considerably lowered, which I consider a positive development.
And no, it's not magic, it's not 'intelligent', it's a predictive algorithm, trained with data built by actual creative and smart human beings.
Just another tool.
-
"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
Wtf is the point of this? Agentic = deletes critical data and I don't need that thanks. My subscription just renewed but I'm going to need to find a new password manager now. -
"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 remember when nerds spent 40 years telling everyone how intelligent they were, and then they were put in charge of everything, and it turns out they're the most gullible dipshits who ever lived
-
"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've used 1Password since V2, advocated for it online & at MUGs, and I find this extremely concerning.
I get it. The VCs want more money; there's the AI hype machine that seemingly makes money fall from the sky.
Come on. 1Password has been the rock to which your users have placed trust in.
Everyone knows AI makes bad code. Programmers under pressure from management have to work faster *will* miss things. This is a fast path to a serious security breech (and of user trust).
-
"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 Bye. Dropping this service immediately. Good job giving me a reason to move my family plan away. Time to check on ChiPass progress.
-
"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 yeah I think it's time to move on from this product. Will be looking at alternatives.
-
"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
If that is the kind of problem you enjoy working on, we are hiring.
abattoir with a sign outside saying "hiring professional farm animals".
-
@abstractcode @1password https://codeberg.org/ChiPass
For a shared option, I have no idea yet…
-
@TeflonTrout @1password AI is a tool. Ignore it at your peril.
-
"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 there are a whole lot of responses to this that show me they didn’t read the article. Thanks for sharing your experiences!
-
"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 bought my first 1Password License in 2009. Upgraded up until 1Password 6. The cloud-forcing and this sloppification are just the nails in the coffin of what once was the objectively best solution.
-
"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 oh so maybe this is why 1Password has been so unstable for multiple colleagues recently...
-
"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 Oh fuck. Not you too!
