Hacker Newsnew | past | comments | ask | show | jobs | submit | causal's commentslogin

"Not everything" - sure, but mass surveillance and autonomous killing are kind of big things to sweep under that rug no?

Output tokens are more expensive

The email use case is interesting, though I feel like dictating via intermediary AI sounds more tedious to me than just using the email app on my phone.

It's not the dictating, it's the clicking and the finding and addressing and the formalities. For a transactional email it's a lot, especially if there's substance to it like needing to click a link and make a choice about some options.

I love the idea of AI agents writing emails that will only ever be read by AI agents.

Thanks for sharing, can you give any specific examples of the kind of personalized tools that you find helpful?

All of the ones I mentioned work great for me. You can dream up your interface and inject technical design ideas and your claw can code it up; as you test it, you fix it if you get new ideas for how it should work.)

For example my recipe project is built around a relatively simple sqlite db. Input for new recipe is in arbitrary format as it goes to an agent (image snapshot on iphone; txt file, speaking it out in a telegram message; a text prompt); the model figures out what to extract and save from the projects’ documents and the context of the specific telegram thread. Output is in whatever form I ask for; rescaled for a number of people; fit in single compact paragraph; or detailed ingredients first then instructions; possibly replace missing ingredients. Query can be arbitrary complex—it almost always finds the recipe I want so far much to my surprise; once I had to followup to select a specific option from a couple ambiguous variants. By default, I like to keep the replies for my recipe queries brief, practical, projected to my cookware. Once I set up this very simple recipe project and imported a couple old datasets, I gave up all of multiple previous recipe schemes (text files or org notes with complex syncing schemes, iPhone apps…)


100%. Every time I see a Claude Code announcement I sigh when I still can't get the desktop app to fullscreen without fritzing or make a plan and display it consistently. Many papercuts.

It's like they oversold capacity and are looking for excuses to get rid of users lol

Opus is fine but it's not THAT much better than the alternatives. I want to support Anthropic because they seem "less shady" than OpenAI but they sure seem determined to push people away.


Claude sets an environment variable to prevent nested invocation, but I've found it can be unset. No idea if they consider that a violation of the secret laws though.

Haven't Github-triggered LLMs already been the source of multiple prompt injection attacks? Seems bad.

Yeah in the span of a month or so we had:

- SDK that allows you to use OAuth authentication!

- Docs updated to say DO NOT USE OAUTH authentication unless authorized! [0]

- Anthropic employee Tweeting "That's not what we meant! It's fine for personal use!" [1]

- An email sent out to everyone saying it's NOT fine do NOT use it [2]

Sigh.

[0] https://code.claude.com/docs/en/agent-sdk/overview#get-start...

[1] https://www.reddit.com/r/ClaudeAI/comments/1r8et0d/update_fr...

[2] https://news.ycombinator.com/item?id=47633396


World: What's your uptime?

Anthropic: 9.

World: 9 nines? That's awesome!

Anthropic: No, a singular 9. 90-something percent uptime.


With mythos, all you need will be 0.9%

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: