HNHacker News
TopNewBestAskShowJobs

foo42

344 karma · joined July 18, 2015

submissionscomments
foo42··on SpaceX says it has agreement to acquire Cursor for $60B
apparently it is a British thing. https://editorsmanual.com/articles/collective-nouns-singular...
foo42··on Sometimes powerful people just do dumb shit
I suspect sycophancy has a lot to do with things . People with power attract those who want their favour and/or money who will align themselves to please the powerful rather than steer them to their best selves. Furthermore as the sycophants accumulate the genuine voices likely begin to sound out of place, like their _against_ the individual. Ironically paranoia gets deployed in the wrong direction to push out the true voices.

I worry this fate will become more common. Everyone can hit up an artificial sycophant at will who they've been told is super intelligent and yet claims their ideas are full of deep insight.

foo42··on LLM Wiki – example of an "idea file"
I built a very similar system into my own assistant type project. In all honesty though I've not used it enough to know how well it works out in practice.
foo42··on POSSE – Publish on your Own Site, Syndicate Elsewhere
I've never implemented it, but I think WebMentions was supposed to enable this
foo42··on Kagi: Small Web Just Got Bigger
Since Kagi small web was on HN a few days ago I've been visiting multiple times a day and spelunking around. I've added a number of interesting feeds to my RSS reader off the back of it already.
foo42··on Astral to Join OpenAI
I was looking into Fedify just yesterday! I'm trying to decide whether to A) try and make my blog an activity pub instance of some sort, B) host my own Mastodon instance, or C) Use someone else's Mastodon server and link to my blog POSSE style. If I go with option A (which somehow feels like how things are _supposed_ to work) Fedify looks like the way to make it happen.

Thanks for your work.

foo42··on Show HN: Three new Kitten TTS models – smallest less than 25MB
minor nit to pick: Welsh accents are British accents as Wales is in Britain. In fact by some definitions it's the most British part.

People from outside the UK often use British as synonymous with English, and in the context of accents, often a South East English accent or some sort of Received Pronunciation (RP) accent. Technically a "British" accent could be from anywhere in England, Scotland, or Wales, and therefore by extension might not even be the English language.

While I'm here, since it's generally confusing, the UK is Great Britain and Northern Ireland. Great Britain is England, Scotland, and Wales.

foo42··on Wander – A tiny, decentralised tool to explore the small web
I feel like activity pub was supposed to / does enable this.

Perhaps someone better informed than I could comment.

foo42··on Many SWE-bench-Passing PRs would not be merged
Yeah I've been wondering if the increasing coding RL is going to draw models towards very short term goals relative to just learning from open source code in the wild
foo42··on Agents that run while I sleep
I've been thinking about this recently and it seems like the most enthusiastic boosters always suggest difference in results is a skill issue, but I feel like there are 4 factors which multiply out to influence how much value someone gets: - The quality of model output for _your particular domain / tech stack_. Models will always do better with languages and libraries they see a lot of than esoteric or proprietary - The degree to which "works" = "good" in your scenario. For a one off script, "works" is all that matters, for a long lived core library, there are other considerations. - The degree to which "works" can be easily (best yet, automatically) verified. - Techniques, existing code cleanliness, documentation etc.

Boosters tend to lay all different experiences at the feet of this last, yet I'd argue the others are equally significant.

On the other hand, if you want to get the best results you can given the first 3 (which are generally out of one's control) then don't presume there's nothing you can do to improve the 4th.

foo42··on Agents that run while I sleep
I salaried employees who are paid by time, and are paying their own Anthropic bills.

Initially there is perhaps a mitigating advantage of briefly impressing ourselves or others with output, but that will quickly fade into the new normal.

Net result: employee paying significant money to produce more, but capturing none of that value.

foo42··on Claude Code, Claude Cowork and Codex #5
> And this is the best country in the world, with the best system of government, because private citizens can voice their disagreement with such actions, including by refusal to participate.

On the off chance other Americans were unaware of this: Other countries are democracies too (and many are better functioning)

foo42··on Ask HN: What Are You Working On? (March 2026)
Sounds fun. Can't beat a good whimsical project!
foo42··on The Eternal Promise: A History of Attempts to Eliminate Programmers
I've been wondering recently if there's some practical path forward for some sort of co-op based LLM training. Something which puts the power in the hands of the users somehow.
foo42··on How will OpenAI compete?
while we're talking pronunciation I'm on an (entirely pointless) one man mission to have "lemon" stick as a pronunciation of "LLM".
foo42··on How will OpenAI compete?
The only two models I ever hear non technical people mention are ChatGPT and occasionally Gemini
foo42··on How will OpenAI compete?
I hope you're right, but is there any guarantee that there will continue to be institutions willing to spend the money to produce open models?

I almost wonder if we need some sort of co-op for training and another for hosted inference

foo42··on NanoClaw moved from Apple Containers to Docker
I've been taking a similar approach with my own exploration.
foo42··on Why I don't think AGI is imminent
I worry that if the reality lives up to investors dreams it will be massively disruptive for society which will lead us down dark paths. On the other hand if it _doesn't_ live up to their dreams, then there is so much invested in that dream financially that it will lead to massive societal disruption when the public is left holding the bag, which will also lead us down dark paths.
foo42··on Building SQLite with a small swarm
I agree.

While I'm generally sympathetic to the idea that humans and LLM creativity is broadly similar (combining ideas absorbed elsewhere in new ways), when we ask for something that already exists it's basically just laundering open source code

foo42··on Using an engineering notebook
Same for me. I also make extensive use of adding links to anything relevant. Spent a bunch of time discussing something in a slack thread: link it. Read some documentation: link it. Had a chat with an llm in a chat window: link it. Writing notes about how a bunch of code works : link to the functions. For this last one I've registered a custom vim:// URL scheme on my system which lets me link to a symbol within a given file, and when clicked focuses the relevant tmux window and navigates the relevant vim instance (using named pipes) to the symbol, or opens a fresh one if not already open.
foo42··on The AI Vampire
After at least a century of labour saving devices being produced and widely adopted in all areas of our lives, how much less time do we spend labouring now?
foo42··on The Physics of Ideas: Reality as a Coordination Problem
I'm currently reading "The disappearance of rituals" by Byun-Chul Han which is highlights the roles of ritual to bring people together into those shared experiences which bind people into a shared narrative and world view.

There are parts of the book I agree with and some I disagree with, but an earlier version of me would have dismissed the whole topic as fluff and any notion of ritual or narrative as superstitious nonsense that needed to be swept away by the light of reason and science. Ultimately though the things that really matter in people's lives tend to be those things which are not coldly rational - love and a place in a wider narrative

foo42··on Go ahead, self-host Postgres
geek code is worthy of its own hn submission
foo42··on 1300 Still Images from the Animated Films of Hayao Miyazaki's Studio Ghibli (2023)
I am in a position I suspect many would be envious of. I've never seen any of these films. My first experience waits ahead!

Where should I begin? I have kids in the 10-15 year old bracket if people think they'd make a good family experience.

foo42··on Blogging in 2025: Screaming into the Void
that's a good analogy
foo42··on Blogging in 2025: Screaming into the Void
Like many here, I too blog for the mental exercise of composing, the human desire to express, and a faith that posting to the public web is some how intrinsically worthy. On the first point I've found I get a lot of benefit from the posts I never even publish as I'll keep chewing over subjects which I'm mentally composing a post about and gain a lot of personal clarity in the process.

One thing I've wondered though (and am mentally composing a post about) is whether there's more good in ai digesting ones writing that we might first feel.

Here's a thought experiment: Would you feel good if someone read your blog and learned something from it? Probably yes. Would you feel good if they passed along something they learned to others, likely in their own words? Probably yes. What if they couldn't recall, or didn't choose to reference where they saw it? Probably still yes, although (speaking personally) my ego would probably prefer they did credit. What if the reader who passed the learning along was the ai?

In a sense we're still contributing to the public discourse and culture when we write, just mediated by models. If a model gives someone a slightly different answer in part because of something you wrote, you've still had an impact on the ultimate human reader.

Just to lay my cards on the table I'm no AI booster, nor doomer. In general I think it's over hyped and may well have a net negative effect if steered by those current at the wheel and consumed without due care, but it has its place where it can be useful.

foo42··on Advent of Code 2025
I find the problem I have is once I get going on a problem I can't shake it out of my head. I end up lying in bed for hours pleading with my brain to let it go if I've not found the time to finish it during the crumbs of discretionary time in the day!
foo42··on Be Like Clippy
Do you have any links to places which would better fit that?

it's perhaps telling of another flaw in this community that I find I need some way to clarify this isn't snark or sarcasm.

foo42··on Writerdeck.org
I was thinking about this subject just this week.

I think an ideal set up for me would be if I could boot my laptop to an different user on the same os where I only had 1 or 2 apps: Obsidian for notes, and possibly an e-reader for some use cases. Then somehow put a timed lock out on the main os/user. That way I could set a period of time where I was locked into that task, but I wouldn't need a dedicated device

← PreviousPage 2 of 4Next →