ArchBeOS
@matta.dev
End to end tests are non-deterministic so why do we write deterministic tests for it?
April 22, 2025 at 1:47 AM
End to end tests are non-deterministic so why do we write deterministic tests for it?
Starting a new project. FOSS LLM hosting for small costs. Privacy focused, easy access, OpenAPI compatibility. It’s gonna be interesting.
March 12, 2025 at 10:15 PM
Starting a new project. FOSS LLM hosting for small costs. Privacy focused, easy access, OpenAPI compatibility. It’s gonna be interesting.
March 5, 2025 at 10:54 PM
Andrej Karpathy is THE authoritative source for AI knowledge for everyone. #ai #dev #software #chatgpt
Deep Dive into LLMs like ChatGPT
YouTube video by Andrej Karpathy
youtu.be
February 5, 2025 at 10:34 PM
Seeing what deepseek has done to the US AI industry is bittersweet. I want the US to succeed but the way it was doing it was not for the good of mankind. Open source is king, and OpenAI will be relegated to a footnote.
January 28, 2025 at 4:37 PM
Seeing what deepseek has done to the US AI industry is bittersweet. I want the US to succeed but the way it was doing it was not for the good of mankind. Open source is king, and OpenAI will be relegated to a footnote.
New #AI Model, Qwen-2.5-1M (in 7b and 14b flavors) is really impressive with its context length of 1 million tokens. Now THAT's a hell of an attention span. This rivals Gemini's context window. Going to think of of ways to use and abuse qwen as a system.
simonwillison.net/2025/Jan/26/...
simonwillison.net/2025/Jan/26/...
Qwen2.5-1M: Deploy Your Own Qwen with Context Length up to 1M Tokens
Very significant new release from Alibaba's Qwen team. Their openly licensed (sometimes Apache 2, sometimes Qwen license, I've had trouble keeping up) Qwen 2.5 LLM previously had an input token …
simonwillison.net
January 26, 2025 at 8:48 PM
New #AI Model, Qwen-2.5-1M (in 7b and 14b flavors) is really impressive with its context length of 1 million tokens. Now THAT's a hell of an attention span. This rivals Gemini's context window. Going to think of of ways to use and abuse qwen as a system.
simonwillison.net/2025/Jan/26/...
simonwillison.net/2025/Jan/26/...
Reposted by ArchBeOS
Speaking of healthcare. Let's make @bsky.app the platform that fights misinformation. Some great follows are @drsusanoliver.bsky.social @jsm2334.bsky.social @therealtruther.bsky.social
Let's get them more followers and engagement than other platforms and let health science accounts find a home
Let's get them more followers and engagement than other platforms and let health science accounts find a home
January 25, 2025 at 5:55 PM
Speaking of healthcare. Let's make @bsky.app the platform that fights misinformation. Some great follows are @drsusanoliver.bsky.social @jsm2334.bsky.social @therealtruther.bsky.social
Let's get them more followers and engagement than other platforms and let health science accounts find a home
Let's get them more followers and engagement than other platforms and let health science accounts find a home
I did a squash instead of a fixup in my git rebase. So annoying when I did this. I feel like an asshole now. Moving to management has really rusted some of my skills... #dev
January 26, 2025 at 4:01 PM
I did a squash instead of a fixup in my git rebase. So annoying when I did this. I feel like an asshole now. Moving to management has really rusted some of my skills... #dev
Stress is a killer and I’m staring it in the face.
January 24, 2025 at 6:08 PM
Stress is a killer and I’m staring it in the face.
Reposted by ArchBeOS
Wow: it seems like Nintendo is beginning to distance themselves from Twitter now, if their newsletter is any indication.
January 19, 2025 at 12:10 PM
Wow: it seems like Nintendo is beginning to distance themselves from Twitter now, if their newsletter is any indication.
This article goes into #DeepSeek and its parent company. Not too detailed, but it shows a hole in the US attempt to bork China's AI ambitions. www.vincentschmalbach.com/deepseek-and...
DeepSeek and the Effects of GPU Export Controls
Last week, DeepSeek unveiled their V3 model, trained on just 2,048 H800 GPUs – a fraction of the hardware used by OpenAI or Meta. DeepSeek claims their model matches or exceeds several benchmarks set…
www.vincentschmalbach.com
January 23, 2025 at 2:10 PM
This article goes into #DeepSeek and its parent company. Not too detailed, but it shows a hole in the US attempt to bork China's AI ambitions. www.vincentschmalbach.com/deepseek-and...
#Bun 1.2 is great, but 90% compatibility for some node std libraries means nothing to me. It might as well be 0% if it isn't a drop-in replacement for legacy node projects.
Bun 1.2
YouTube video by Bun
youtu.be
January 21, 2025 at 10:13 PM
#Bun 1.2 is great, but 90% compatibility for some node std libraries means nothing to me. It might as well be 0% if it isn't a drop-in replacement for legacy node projects.
Android Programming feels so much like ASP.NET post backs from 2004.
January 21, 2025 at 9:15 PM
Android Programming feels so much like ASP.NET post backs from 2004.
#DeepSeek-R1 is a game changer. FOSS thinking models are the future. Weird to say, but thanks China?
January 21, 2025 at 8:26 PM
#DeepSeek-R1 is a game changer. FOSS thinking models are the future. Weird to say, but thanks China?
The future of #programming is #acceptancetesting
Acceptance Testing Is the FUTURE of Programming
YouTube video by Continuous Delivery
www.youtube.com
January 15, 2025 at 9:00 PM
The future of #programming is #acceptancetesting
Silent Hill 2 has to be set in the US on account of all the handgun ammo I keep finding lying around.
January 4, 2025 at 1:16 PM
Silent Hill 2 has to be set in the US on account of all the handgun ammo I keep finding lying around.