blfr

3 ms·

1.
by blfr 5d ago
This is a story of enshittification but... I don't see any of it and neither should anyone who reads HN: * adblocking is simple both in browser and on the dns level, * revanced youtube allows you to even skip sponsored segments, regula
2.
by blfr 7d ago
They're often directly applicable: minor bugs, inconsistencies, missing tests, commonly also stuff I already now (like some infra config details).
3.
by blfr 7d ago
I do neither maths nor science with AI but in my experience most models are perfectly willing to burn tokens on a ton of sidequests at the earliest opportunity.
4.
by blfr 7d ago
It's just so nice when even crazy people agree to some sane standard. Like USB c, like having dark mode, just nice.
5.
by blfr 8d ago
Yes but I use separate passwords and webauthn is per domain as well.
6.
by blfr 8d ago
The creds they store hold no value. These creds only give access to their system anyway. They store the email address either way so the situation is no better but less convenient.
7.
by blfr 8d ago
So are companies, right?
8.
by blfr 8d ago
Email 2fa otoh is the worst UX I have experienced. I curse every time Claude sends me a magic link. Absolutely hate them.
9.
by blfr 8d ago
Other companies have no qualms about distilling the first. Let's hop on gear and get the market to deliver a distilled Fable that runs on a smartwatch. Sooner is better.
10.
by blfr 8d ago
Yes, thank you. Maybe I am getting old but password+yubikey/webauthn was really top UX.
11.
by blfr 8d ago
All IT infrastructure is a problem waiting to happen. I have backups and a rollback mechanism.
12.
by blfr 8d ago
Why?
13.
by blfr 8d ago
Oh, yeah. I gave Claude sudo on my AI dev machine: it commits, pushes, checks the results of pre-push tests, sudos to another user, deploys, checks the results of smoke tests, and then checks the end result. It also has MCP access to Cloudf
14.
by blfr 12d ago
I use Twitter daily and haven't noticed any screws. How is Elon supposedly turning them?
15.
by blfr 12d ago
Mongols did it on horseback. It's mostly a matter of will/capability and no scruples.
16.
by blfr 12d ago
The fondness for TEEs is mostly Moxie's and I think he's not involved with Signal anymore. But he does have an AI chat/inference based on enclaves! https://confer.to/blog/2026/09/confidential-wo
17.
by blfr 12d ago
Yes, very much so. It is basically a standard across the western world for politicians, journalists, whatsapp refugees... Of course there are many alternatives but most of them have most of their users here, on HN. MacOS, iOS, Windows, Linu
18.
by blfr 12d ago
Maybe it's because I use Molly as a secondary device (my tablet) but I never had an issue where it didn't work for weeks.
19.
by blfr 12d ago
I meant the extinction part not software getting better.
20.
by blfr 12d ago
OP's central rebuttal is that anxiety or fear are not evidence. This particular anxiety has been with us for thirty+ years, its proponents/sufferers weren't right before, and they don't have some new evidence now.
21.
by blfr 13d ago
Hacking model is the aligned model. I don't like it when the model refuses to sidestep some throttling limit or scan my own codebase for security issues. I want full-on exploits in my test suite. With LLMs the code going to prod should
22.
by blfr 13d ago
Yes, benchmarks are gamed and only loosely indicative of real world performance. Also yes, Fable is massively better than Opus. It requires significantly less instruction and specs and produces more directly mergeable code. The improvement
23.
by blfr 14d ago
I have considered it, I've been hearing way more of it that I like and it's rationalist slop with little to no predictive power: https://foom.hyperplex.org/
24.
by blfr 14d ago
I enjoy Claude Code very much, have max privately and team premium at work, but the doomer marketing and this whole regulate-while-we're-ahead spiel is extremely annoying and makes me wish Anthropic gets trounced.
25.
by blfr 16d ago
Yes, I do the same with Claude Code. Create an instance on the server for a project and then can create sessions from any device, close my laptop while claude code keeps working, etc. without losing the convenience of dedicated apps.
26.
by blfr 17d ago
If you ask them beforehand, you get a self-selecting sample, biasing your result. You get better results by not asking them.
27.
by blfr 17d ago
Hilariously, there is no substitute for intelligence, not even another intelligence.
28.
by blfr 17d ago
Humans made Stuxnet and nuclear weapons.
29.
by blfr 17d ago
The core reason to enforce trademarks is not to fluff the corporation that holds it but to let the public be confident in the name. So the trademark should hold for as long as a reasonable person can be confused by the new Twitter regardles
30.
by blfr 17d ago
As a Linux user of over 20 years, I really don't like thinking of myself as some Aborigine barely holding on to a scrap of ancestral land. My tech preferences have become so ubiquitous that nearly the entire Internet runs them. In Ang

More ›