tao_oat

7 ms·

61.
by tao_oat 6mo ago
Effective Altruism: https://en.wikipedia.org/wiki/Effective_altruism
62.
by tao_oat 6mo ago
Was not meant ironically. I understand how in practice it's a political role (as most positions of power are) but I think it's applying the wrong standard to expect religious leaders to communicate like politicians.
63.
by tao_oat 6mo ago
And a few paragraphs before: > And then, and I’m sorry to be so blunt, then it’s die or kill.
64.
by tao_oat 6mo ago
The author seems to have some cognitive dissonance. For a piece saying that you cannot justify violence, there sure seems to be an awful lot of justifying violence in here.
65.
by tao_oat 6mo ago
The Pope is not a politician.
66.
by tao_oat 6mo ago
https://archive.is/c2sBh
67.
by tao_oat 6mo ago
> The current tech is a sigmoid What makes you say that? I'm only asking because the data I've seen looks pretty cleanly exponential still, e.g. https://metr.org .
68.
by tao_oat 6mo ago
> Our tests gave models the vulnerable function directly, often with contextual hints (e.g., "consider wraparound behavior"). "Often with contextual hints" is doing some heavy lifting here, IMO. I agree with the artic
69.
by tao_oat 6mo ago
The page says > I recommend printing these toys on HEAVY CARD STOCK.
70.
by tao_oat 6mo ago
This is a really great overview; what a useful resource for other open-source projects.
71.
by tao_oat 6mo ago
Those products have very low COGS in comparison to this.
72.
by tao_oat 6mo ago
A bit surprised by the snarky comments here -- I also want Claude to work reliably but very few (no?) companies have ever seen this level of rapid growth. We're going to go through a long fail-whale-style period and I can imagine very,
73.
by tao_oat 6mo ago
See also: https://github.com/Barre/ZeroFS
74.
by tao_oat 6mo ago
Unfortunately this has all the hallmarks of AI writing, which made me a lot less motivated to read it.
75.
by tao_oat 6mo ago
I tried to move my sites to Bunny Edge Scripting and found the experience mostly poor, unfortunately. A lot of failures without error logs, and purging the pull zone cache only seemed to work sometimes. A shame because I like their offering
76.
by tao_oat 7mo ago
Relevant: https://days-since-openclaw-cve.com/ Currently we're at 1.8 CVEs per day since OpenClaw launched!
77.
by tao_oat 7mo ago
Overall I think it's useful. Superpowers has several skills. Its core workflow is: - Brainstorm with you to design a spec - Use subagents to review its own spec, then get your approval - Based on the spec, write a plan, use subagents t
78.
by tao_oat 7mo ago
I didn't find this very convincing. Especially the argument around artificially low cost -- we know that training the next model is the biggest cost for these companies, and we've already seen inference costs fall drastically ( ht
79.
Just One More Prompt(btao.org)
2 points by tao_oat 7mo ago|0 comments
80.
by tao_oat 7mo ago
This feels somewhat ahistorical. The US has nearly always been successful in terms of conventional firepower and individual operations. E.g. in 2003 the US overthrew Saddam's government in a matter of weeks. The US won most battles in
81.
Show HN: What % of your commits were written by AI?(technically-your-name-is-on-it.btao.org)
5 points by tao_oat 8mo ago|2 comments
82.
by tao_oat 8mo ago
Haven't tried them in enough depth to compare. Nanobot's was not great (cron + a HEARTBEAT.md meant two ways to do things, which would confuse the AI). But because the implementation is so simple, I could improve it in a few minut
83.
by tao_oat 8mo ago
This is an interesting idea! I searched around and it looks like there's [ast-grep]( https://ast-grep.github.io/ ), an AST-aware CLI that can search and refactor code -- and you can expose it to your AI agent using a ski
84.
by tao_oat 8mo ago
And they love to do this in spite of writing "NO FALLBACKS" etc. in your AGENTS.md.
85.
by tao_oat 8mo ago
I haven't used them all but based on my partial research so far: - OpenClaw: the big one, but extremely messy codebase and deployment - NanoClaw: simple, main selling point is that agents spawn their own containers. Personally I don&#x
86.
by tao_oat 8mo ago
I'd apply to work for Anthropic in a heartbeat if it was a European company.
87.
by tao_oat 11mo ago
I've been trying to get ChatGPT to stop adding this kind of fluff to its responses through custom instructions, but to no avail! It's one of the more frustrating parts of it, IMO.
88.
How to hire the best people you've ever worked with (2007)(fictivekin.github.io)
1 points by tao_oat 1y ago|0 comments
89.
by tao_oat 1y ago
neat landing page but i don't see how their distribution model would be fundamentally different from / independent of the app stores.
90.
by tao_oat 1y ago
Skills were released in Claude Code, what, yesterday? I doubt there's a simple answer to this -- it'll depend on the model, task, etc. You could try to get your agent to test its own skills. From https://blog.fsck.com&#

More ›