Nothing here yet.
Nothing here yet.
Mar 22 · 3 min read · I spent the last few weeks deep in the weeds of AI limitations. Jailbreaks, inference quirks, why models hallucinate when pushed too far. Core DL stuff. And somewhere between trying to break LLMs and
Join discussion
Jan 25 · 3 min read · I recently built a small web project using GitHub Copilot AI Agents. It took about 15–20 minutes, maybe 2–3 prompts, and most of the visible structure was done. Let me get this out of the way early:this is not an achievement. What is worth writing ab...
Join discussion
Jan 1 · 6 min read · Meta’s AI strategy in 2025 is no accident. It follows a two-pronged playbook: own the data & pipeline that make models reliably useful, and own the agentic interfaces that turn model outputs into repeatable user habits. The Scale AI move sealed a ver...
Join discussion
Dec 26, 2025 · 3 min read · When people talk about big tech acquisitions, they usually imagine clean headlines. Company A acquires Company B for X billion dollars. Regulators react. Antitrust lawyers wake up. The story is easy to tell. What Meta did with Scale AI does not fit t...
Join discussionDec 26, 2025 · 3 min read · A few days ago, I took a clear position. I argued that purpose-built AI architectures like LPUs and wafer-scale systems would eventually outcompete GPUs. From that, I went a step further and predicted that NVIDIA might not be the ultimate winner of t...
Join discussion