CTO at @speechblubs · Developer · Speaker · Mentor · iOS · Machine Learning · AI · Vision
Slovenia
Joined February 2011
- Tweets5.1K
- Following485
- Followers345
- Likes984
6.1 Terra really does not perform bad in visual reasoning. New test on LegoBench: legobench.net/runs/20261027-…
Week 3: So it's not exactly an open source version of the existing software, but it's an open source version of patches you need to run to get Skyrim running on Apple Silicon. Details on my blog: unifiedsense.com/blog/skyrim…
Interesting how benchmaxxed models perform bad on LegoBench, how are they outperforming Opus and Astra on almost all benchmarks, but fail so horribly on something so simple as putting bricks together? AGI is here.
Kimi K3.1 Pro barely has a clue.
legobench.net/runs/20261023-…
In addition to all of the other ports, Skyrim also runs natively on Apple Silicon. iPad, iPhone and Mac. :)
For my slovenian followers, I made a couple of videos explaining AI concepts. First one is about AI Agents. Let me know your feedback!
youtube.com/watch?v=TDmxmT5E…
A little bit late, but still - week 2 in Open Source: My version of Adobe Creative Suite. This would be a much more exciting if ArtCraft wouldn't have been released few weeks ago. Tauri 2, Rust and WASM engine. Covering most of their apps and a little more.
unifiedsense.com/blog/monet-…
I had used up my Codex Pro 500 plan yesterday to 100%. Just now I noticed I still have 20% left. Did anyone else notice this?
Dal Rupnik retweeted
We now have state-of-the-art pre-trained base models in the 3T-parameter weight class, built entirely in-house at MAI.
Post-training results coming soon -- stay tuned!
Lots of details in our technical blog post.
microsoft.ai/news/frontier-p…
Were you ever wondering how good is AI at building Lego models? I built a simple benchmark. GPT 6 Astra is ahead. Blog post here: unifiedsense.com/blog/legobe…
I'm thinking that commercial software ends in 2027-2028. Thoughts?
unifiedsense.com/blog/end-of…
I built OpenGlaido, open-source dictation for macOS, Windows, and Linux. Bring your own model, or run one locally on a Mac.
unifiedsense.com/blog/opengl…
openglaido.com
Dal Rupnik retweeted
Meet Mistral Large 4, aka Le Chonk.
• 1T parameters, natively multimodal. 49B active.
It is the best open weights model from US or Europe on aggregated benchmarks.
• State-of-the-art on critical workloads, including cyber defense, manufacturing and finance and it surpasses closed frontier models on visual grounding.
• Forged in Europe end-to-end and is deployable from Europe via our own Mistral Cloud infrastructure.
• Available to all via API today. Working with cybersecurity partners privately.
Open weights release end of October.
🤖 Made with AI
Dal Rupnik retweeted
GEMINI 4 ARGON IS ROLLING OUT TO ULTRA USERS? 👀
A friend just sent me this screenshot, Gemini 4 Argon randomly appeared in his Antigravity model picker, and he said he actually used it for a while..
is Google quietly rolling it out already??
as they already said Ultra users will be among the first to get Argon access, but I haven't seen an official Antigravity rollout yet..
Anyone else seeing this, or is he messing with me? 😭
Did anyone notice that when upgrading from Codex Pro 200 to 500, you do not really get a quota increase? I spent a day with Codex 200 and then once used, I updated to 500. I continued to work on the same projects. Quota decreases exactly the same.
