avi retweeted
my *personal* opinion about this is that it's basically too late w/ deepseek flash 1017 and we should just rip the bandaid off and let everyone rip everything apart together
we need to trust in humanity and give them the tools to defend themselves, instead of baseline expect criminality
avi retweeted
"whatever you now find weird, ugly, uncomfortable and nasty about a new medium will surely become its signature. CD distortion, the jitteriness of digital video, the crap sound of 8-bit - all of these will be cherished and emulated as soon as they can be avoided."
- brian eno
This is such a banger rule in @opencode's AGENTS.md
"Before adding complexity for a speculative or vanishingly unlikely race or security edge case, explain the concrete failure mode, likelihood, and complexity cost to the user and get their buy-in. Do not silently expand scope for theoretical robustness."
avi retweeted
I kid you not, sometimes I talk to an AI and they are stunned by my insight into architecture
avi retweeted
There must exists an app with PHP, deployed on Railway, with calls to cURL, with other services written in node that use Brotli compression running on a couple day old Linux, with their code self managed on GitLab
They are having a hell of a week
avi retweeted
DeepSeek V4.1 Flash 1017 runs at 1,521+ tok/s on Fireworks Ultrafast.
Fast enough to ship Tetris in 11 seconds, 7x faster than Claude Opus 5.5.
Join the Ultrafast waitlist: bit.ly/3UA0wKq
avi retweeted
I just realized that @EffectTS_ would probably be essentially ignored if released today even in the exact same state
It’s just so big that everyone would assume it’s slop and there’s no way someone actually put enough thought into making something that big actually well designed
avi retweeted
🚀 Introducing DeepInfer, our in-house inference framework, now fully open source!
Powering DeepSeek-V4.1-Flash at scale with high throughput and low latency. We’re sharing our production stack to help the community lower compute costs, reduce energy use, and make AI more affordable.
🔗 github.com/deepseek-ai/DeepI…
1/4
avi retweeted
Introducing Motherload — import data files from TypeScript with full type information
Powered by TypeScript 7.1's content mappers (now in beta)
.md
.toml
.yaml
.csv / .tsv
.jsonc / .json5
.txt
.env
We're truly living in amazing and terrifying times.
In the past few months, I've built:
‣ A Netflix/YouTube alternative for my son: I upload the exact videos I want from any video platform just by entering the URL. He gets a nice experience with no algorithms.
‣ A writing tool that helps me with grammar, fact-checking, alt text, etc.
‣ A visual regression tool that can take hundreds of thousands of screenshots for $5/mo (compared to thousands of dollars with other solutions).
‣ A browser extension that lets me pin comments to any element on any page so I can generate a prompt for an AI agent.
‣ Lots of fun educational mini-games for my son whenever he has trouble learning something.
And plenty more.
I wouldn't have had time to build any of this without AI.
avi retweeted
Skeletons are the worst cancer in the history of user interfaces. Just make loads fast. Ping across the ocean is never more than 200ms, maybe 300ms if you're Australian, just return the damn data at that speed and you never need to even think about this crap.
avi retweeted
I was interviewing someone with 4 years of OCaml experience, and said:
"The language solved problems I didn't know I had, but the tooling invented problems I didn't know were possible"
avi retweeted
How long it took me to find each SWE job:
2014: 1 job applied, 0 interviews, 1 day (nepotism)
2015: 1 job applied, 2 interviews, 1 week
2016: 1 job applied, 4 interviews, 1 week
2016: 1 job applied, 0 interviews, 1 day (friend is a CEO)
2018: 1 job applied, 2 interviews, 1 week
2019: 1 job applied, 5 interviews, 1 week
2021: 1 job applied, 0 interviews, 1 day (invited)
2022: 7 jobs applied, 16 interviews, 3 months
2026: 167 jobs applied, 54 interviews, 1 year 4 months
avi retweeted
Saw a guy call it cuck coding instead of vibe coding because you just sit in a chair and watch😭
in 3 weeks we're gonna have @DarioAmodei announce that @AnthropicAI had an even bigger more dangerous scarier spy caught in their office. He will have done more than just place a bug or two.
We discovered a sophisticated recording device that was hidden inside one of our offices. We used our models to understand the device and where it had come from, identifying that a specific employee had planted it a few days earlier. We referred the matter to the FBI, and we’re grateful to the agents and prosecutors who moved quickly to make an arrest. The FBI’s investigation is ongoing and we will continue to support their work.
At this time, we have no indications of this individual inappropriately accessing or exfiltrating model weights, source code, or customer data. The employee was terminated.
We take protecting our technology, data, and people seriously, and as a matter of national importance. Our security program is designed to protect against a wide range of threat actors, from insider risks to cybercriminals to nation states, and this case is an example of how AI can help address these sophisticated security threats.
GLM-Legal uncensored:
Introducing GLM-Legal
Built on GLM-5.3-Flash, GLM-Legal matches frontier models on mainstream legal benchmarks.
Designed for real legal workflows: contract review, evidence analysis, compliance, and more.
Tech Blog: z.ai/blog/glm-legal
Apply for API Access: z.ai/consultation/legal
Try AutoClaw: autoclaw.z.ai
avi retweeted
Good idea
(Building something to make this even better...)




























