Ted L
@Zaneris
Followers
112
Following
10K
Media
51
Statuses
2K
Co-Founder, .NET Developer, RCAF Veteran, @_zaneris (Threads)
Toronto, Canada
Joined June 2008
The @karpathy interview 0:00:00 – AGI is still a decade away 0:30:33 – LLM cognitive deficits 0:40:53 – RL is terrible 0:50:26 – How do humans learn? 1:07:13 – AGI will blend into 2% GDP growth 1:18:24 – ASI 1:33:38 – Evolution of intelligence & culture 1:43:43 - Why self
520
3K
18K
Excited to release new repo: nanochat! (it's among the most unhinged I've written). Unlike my earlier similar repo nanoGPT which only covered pretraining, nanochat is a minimal, from scratch, full-stack training/inference pipeline of a simple ChatGPT clone in a single,
663
3K
24K
Today, we’re announcing a multi-year, multi generation strategic partnership with @OpenAI that puts AMD compute at the center of the global AI infrastructure buildout. ✅ 6GW of AI infrastructure ✅ Initial 1GW deployment of AMD Instinct MI450 series GPU capacity beginning 2H
291
1K
7K
EVERYONE has been asking, so here it is. The first Ubiquiti video on my new channel... and I managed to break it. oops. https://t.co/NVm3hV6wW2
12
21
846
Introducing Claude Sonnet 4.5—the best coding model in the world. It's the strongest model for building complex agents. It's the best model at using computers. And it shows substantial gains on tests of reasoning and math.
1K
3K
21K
Our data says that 50/50 or slightly unequal splits has basically no bearing on rate of success. Very unequal splits for a two founder team (like 80/20, etc) are negatively correlated with success.
50/50 co-founder splits are the signal of a weak CEO
23
9
338
So here’s the story behind why yesterdays live #metaconnect demo failed - when the chef said “Hey Meta start Live AI” it activated everyone’s Meta AI in the room at once and effectively DDOS’d our servers 🤣 That’s what we get for doing it live!
401
484
12K
Meta Connect 2025 is here! Watch the keynote livestream tonight @ 5 pm PT to see the future of AI wearables and beyond: https://t.co/VygfGlE1Fv
40
50
266
Just what everyone needs, software from alibaba running directly on their machine… run their models locally through ollama or LM studio.
🚀 Qwen Chat Desktop for Windows is here! 💻 All the power of Qwen Chat — now with MCP support for smarter, faster agents. ⚡ Run up MCP Servers, supercharge your productivity, and stay in control. 📥 Download now → https://t.co/uYQIIGQAJo
0
0
1
Looks like ML vision is nearly a solved problem
Introducing DINOv3: a state-of-the-art computer vision model trained with self-supervised learning (SSL) that produces powerful, high-resolution image features. For the first time, a single frozen vision backbone outperforms specialized solutions on multiple long-standing dense
0
0
2
GPT-5 just finished Pokémon Red! 6,470 steps vs. 18,184 for o3! Check the stats site to compare! That's a huge improvement! Well done, @OpenAI you cooked with GPT-5. What an incredible model. Next up: GPT-5 vs. Pokémon Crystal (16 Badges + Red). The run starts soon on Twitch.
GPT-5 has reached Victory Road! This is the last challenge before the Elite Four. GPT-5 reached this part almost three times faster than o3 (6105 steps for GPT-5 vs 16882 steps for o3). Here are my observations as to why: - GPT-5 hallucinates far less than o3. This is the main
28
93
654
gemini-cli still isn’t caught up to Claude Code but it’s nice to see the competition
Gemini CLI now integrates into VS Code. Get intelligent, context-aware suggestions and native in-editor diffing to review and modify changes directly in your IDE. Learn more: https://t.co/avOsSn4Enp
0
0
0
Meet Gemini CLI GitHub Actions, now in beta ✨ Your no-cost AI coding teammate to help automate tasks, review code, and ship faster → https://t.co/rj2rVFxJWw
19
91
743
You can use /model sonnet[1m] to switch to the 1M context version of Sonnet 4 in Claude Code* *only works via API auth though, not on a Pro/Max subscription props to @AlexGrama for the tip
28
46
491
This is great news for Claude Code, it’ll be interesting to see how the intelligence holds up as you approach the context limit.
Claude Sonnet 4 now supports 1 million tokens of context on the Anthropic API—a 5x increase. Process over 75,000 lines of code or hundreds of documents in a single request.
0
0
0
Perplexity has offered to buy Google Chrome for $34.5 billion. That’s even higher than Perplexity’s own valuation
theverge.com
The unsolicited offer is higher than Perplexity’s valuation.
23
31
380
New image gen model coming along with Gemini 3 it looks like. Squadron on their way to take out #GPT5
0
0
0