Joined October 2012
- Tweets4.4K
- Following4.6K
- Followers1.9K
- Likes74.9K
the hyperion cantos is still one of my very favorite book series. highly rec all four books! then read accelerando by charles stross
aulneau retweeted
Decision models could damage Anthropic's business.
You don't need Opus 5.5 when you can instantaneously run hundreds of natural language assertions on agent output...
... at a fraction of the cost.
This approach is being streamlined via Traits.
traits.sh/
Hey @simonsarris, I love nocta, works great on my pixel. Have you thought about making it so we can create custom presets? I want to try and make a ricoh preset 👼 people could also share presets
aulneau retweeted
Hi, I'm Felix. I'm an eng lead working on Claude. I love watching people take flight and do things they never thought possible.
We often describe that as "being more ambitious" when using AI, which is tricky advice to follow. I wrote a tutorial:
felixrieseberg.com/how-to-be…
aulneau retweeted
OpenCode 2.0
anoma.ly/notes/opencode-2-0/
aulneau retweeted
Benefits of the CAFE stack (@Cloudflare + @alchemy_run + @foldkit + @EffectTS_) for building software with coding agents:
1. Agents invent less architecture. They focus on product-level execution within a well-defined architecture. You build faster, with more confidence. You focus on what makes your product unique.
2. Your codebase becomes semantically rich. Intent and meaning are explicit rather than left for future agents to infer. Each new agent spends fewer tokens reconstructing intent.
3. Tests read like executable behavioral specs. Expected behavior is easy to inspect, and both humans and agents can verify implementations against it.
4. Strong conventions result in remarkably legible AI-generated code. Agents do what you expect more often, with less steering. The codebase feels consistent and predictable.
5. Explicitness makes behavior clear. Unexpected magic has fewer places to hide. Diffs are easier to review. Yes, it’s more LoC, but who cares when you’re not writing it by hand?
aulneau retweeted
we specifically avoided using quickjs as our codemode implementation
it brings in webasm complexity + forces worker threads + serialization of values back and forth
if you're involved in implementing something like this check out our codemode package it is way simpler
aulneau retweeted
the assembled lore for idiomatic effect project organization ratstack.sh/lore/structure-e…
aulneau retweeted
[email protected] brings a ton of new stuff, including broad GCP coverage, Neon becomes more than a Database, pluggable secret providers, Prisma ORM, better scheduling primitives for Durable Objects and much more.
alchemy.run/blog/2026-09-29-…
aulneau retweeted
My hot take from chatting to @poteto is that we should use MORE abstractions in the AI age
You can use them (combined with harsh lint rules) to reduce the design space available to the agent and constrain them only to good decisions.
Combined with the fact that high-leverage abstractions let you do more with less code - so, more token efficient.
Plus, unwinding the damage from a bad abstraction is much cheaper with agents.
This runs counter to a lot of folks thinking that agents just want to read the raw code. They can, but they're not maximally efficient that way.
Be braver! Design abstractions.
aulneau retweeted
Cloudflare have integrated Pi Durable in the agents SDK. Really happy how it turned out. Props to @mattzcarey et al. for getting this done on such short notice.
It "just works" :)
github.com/cloudflare/agents…
aulneau retweeted
i think i'm gonna frame this one :)
Mods are absolutely insane. You can now customize Claude to work and look the way you want by just prompting it.
Each person works differently, so there's no reason why everyone should have an identical Claude experience. Make Claude your own, and share mods as plugins so others can try your mods too.
aulneau retweeted
i'm beyond excited for what can be built with this
People of Pi: We've shipped Pi 1.0 with Pi Durable. Go make them yours.
earendil.com/posts/pi-1-0/
