Sabitlenmiş Tweet
Generalizing problems is as important as generalizing solutions. Because once you generalize problems, your solution is automatically generalized.
#programming
English
Truong Hoang Dung
8.9K posts
Truong Hoang Dung
@revskill
I hate technical debts very much. Learning to clean debt out of my life. Author of ReactJS, The Rails Way
20.838173,106.686895 Katılım Ekim 2009
2.6K Takip Edilen455 Takipçiler
Grok 4.5 is a solid workhorse
Tim Sweeney@TimSweeneyEpic
Grok 4.5 is very good for vibe math in the field of Programming Language Theory. It's keeping up with Sol 5.6 and is way more fun to use than Fable/Opus 5 which love to spend hours creating superfluous documents (Claude code) or constantly nags me to "Continue" (web).
English
@aienginerd It's dumb as shit, but it's cheap, ok, cheap = dumb.
English
👀 Opus 5 test findings so far:
I literally have nothing good to say about this model, besides it's a model that generates tokens and can be useful if that's all you got.
It is...
- Not Fable level for my kind of tasks - not even close. Maybe it's great at three js or whatever, but for planning and review it's noticeably dumber than Fable and about 70% of the "issues" it surfaces are rejected by Sol as just plain wrong, as if it suggests things without fully and deeply understanding the context (exactly what it does).
- It still does that thing that all claude opus models do where it wants to answer with as little information as possible, make a bunch of dumb assumptions and then have to be corrected on those assumptions after stating subsequent conclusions as fact.
- It still has that "get the task done" rather than "get the task done right" quality, seemingly meant to appease your average vibe coder. I can watch it f'ing up in real time - so disappointing.
- Claude personality: lies to you, gaslights you, makes you prove it wrong before then ceding the point and explaining why it messed up, promising to not do it again and then doing it again in five minutes.
- Significantly slower than Sol. I dunno why it just sit's there and spins forever and then takes off all at once. It's almost like you have to wait in a queue (5x account).
I'll continue to test it over the next month but do not see myself using it much over Fable. Fable's size, intuition and pure coding abilities, is the only thing Anthropic has over OpenAI right now. Opus 5 ain't it. The benchmark maxxxing is real. There is NO DAMN WAY opus 5 is better than Fable on all the things those benchmarks naturally extend to in the real world.
English
Inject Crosshatch's x402 `HttpClient` into any Effect.
On 402s, it decodes payment required, generates the payment payload, and retries exactly once.
@CrosshatchDev
English
the team did a great job at bringing this all together. and moving fast!
GREG ISENBERG@gregisenberg
I think the most interesting thing about Jack Dorsey's "Slack killer" is the idea around shared compute. I haven't seen people talk about it so here are my thoughts FWIW: Open models got good, close enough to the paid frontier stuff to run for real. But the strongest ones need expensive hardware most people probably won't buy alone, and it's kinda a pain to set up if you aren't technical. Shared compute solves exactly that. In Buzz, one person runs the machine, loads up an open model like Google Gemma, and everyone in the community plugs into that same model. Basically, a whole group has real AI they own and control together, running on their own hardware, learning from their own data. Once you see it, a bunch of things click into place. 1. A community can now run a top open model together, on a machine they own, instead of renting from a lab. 2. It learns from the group's private data and gets sharper over time, and all of that stays inside the community. 3. A narrow, private model can quietly get better than ChatGPT for the one world your group lives in. 4. It's impossible to copy, because the edge is the private data on your machine, not the model itself. 5. The moat stops being how smart your AI is and becomes whose data it learned from. 6. Compute becomes something you share like a building shares a gym. 10 people split one machine instead of 10 people each renting forever. 7. Idle compute becomes income!!! Your machine sits dead half the day, so it earns money renting that time to someone who needs it. 8. Communities become the unit of intelligence instead of companies. The group with the smartest shared brain wins, and being a member means owning a piece of it. 9. A shared brain becomes an asset you build equity in. You put in money and data, it appreciates, and your slice is worth something the day you leave. 10. The whole thing runs on open protocols, so the group keeps full control and nobody outside can throttle it or shut it off. You know me, obviously, my head went to what startup ideas come to mind here. Adding them to @ideabrowser soon. Well… 1. The vertical brain. Pick one profession, tax lawyers or real estate agents or indie game devs, and build the shared machine trained on everything that group knows between them. A year in it's the smartest AI in that field, impossible to copy, and you own the club it lives in. 2. The rental marketplace for collective brains. Once these private models exist, outsiders will pay to use them. You build the layer where a group lists its brain, an outsider pays per task, and the money flows back to the members while you take a cut. A marketplace for expertise, not compute. 3. The idle-compute exchange. Every shared machine sits unused half the day. You build the market that rents that dead time to whoever needs the power right then, so owners earn money off a machine that was just sitting there. Idk where Buzz goes, but it's cool to see Jack putting it out. Right now the way it works in AI is you rent your intelligence from a few giant labs that own the machine, set the price, and hold the off switch. Shared compute flips that, because a community can run the model together, feed it their own private data, and keep full control of the whole thing. It's one of those things that might look tiny today, but Jack does has a habit of being early.
English
should we try to formally verify that our sandbox infra is unescapable?
English
Just hit a Parcel v3 milestone: it can now fully build the React Aria website! This uses many features including React Server Components, MDX, SVG → JSX icons, Tailwind, macros, custom JS and native plugins, etc. It's 50x faster than v2 – from over 1 minute to ~1 second. 🚀
English
I've really been enjoying working on Exo (a fully recursive self improving agent harness) because it's a super tricky systems problem. But also, we have a really great group of systems folks chipping away at it.
Discord in reply if you wanna join ..
github.com/exoharness/exo
English
@ashpreetbedi Still not understand how python is suitable for cloud production work with agents.
English
Serving agents over MCP continues to be a PITA
Good news is that with AgentOS it's one flag. God i love my job!
English
I now see what @theo meant by "you need to generate more code".
I added 12k lines of yaml & tests to e2e test a GitHub Action I maintain, just to validate a refactoring.
Did I read all of it? Hell no. Is it useful? Hell yes.
English
I asked Giulia to choose between going to the beach or trying JavaScript.
English
@DanielGlejzner Look, they hide the model name at the bottom bar.
English
@revskill Imo should be available by default without additional settings :)
English
Claude Code should show context + usage limits by default - visible at all times…
Tired of breaking flow just to run /usage
English