This is the third in my annual series reviewing everything that happened in the LLM space over the past 12 months. For previous years see Stuff we figured out about AI in 2023 and Things we learned about LLMs in 2024.
It’s been a year filled with a lot of different trends. — Read More
Daily Archives: January 1, 2026
Cybersecurity Changes I Expect in 2026
It becomes very clear that the primary security question for a company is how good their attackers’ ai is vs. their own.
— ISOs increasingly realize that there is no way to scale their human team to deal with how constant, continuous, and increasingly effective their attackers are becoming at attacking them
— It becomes a competition with how fast you can perform asset management, attack surface management, and vulnerability management on your company, but especially on your perimeter (which includes email and phishing/social engineering)
Read More
Planetary-Scale Deep Reasoning: Building Our Final Presidential Daily Brief Prompt & Comparing Gemini 3/2.5 Pro/Flash ASR/TOC
Over the last few days we have been exploring having Gemini 3 Pro “watch” an entire day of television news from a given channel from across the world and write a deeply reasoned and researched intelligence-style report that looks across all of that coverage and teases out the overarching themes, narratives, implications and future impacts of the day’s events. Yesterday we had Gemini 3 Pro interactively improve its own prompt to generate a final “ultimate” prompt to write a Presidential Daily Brief (PDB)-style intelligence report from a day’s broadcast transcripts. Today we’ll add a few final refinements and then demonstrate our new prompt on a single day of a Russian television news channel across Gemini 3 Pro, Gemini 3 Flash, Gemini 2.5 Pro and Gemini 2.5 Flash Thinking using both the full-day Chirp 1 ASR transcripts and a preprocessed story table of contents. No data was used to train or tune any model. — Read More
China Just Pulled Its Own Manhattan Project and No One Saw It Coming
Or: The West banned the machines. China hired the machinists. Sometimes plans just do not go how you planned them. Ironically, I had been writing this article for a month now, and all research pointed at China being way too far behind. Well…
December 2025. Reuters reveals that China completed an operational EUV lithography prototype in a high-security Shenzhen facility. Not through reverse engineering captured ASML machines. Not through some breakthrough in domestic optics manufacturing. Through something far simpler.
They recruited the humans who knew how to build them. — Read More
Rick's Cafe AI 