[Updated 6th January 2026] - things I am thinking about and discussing over coffee this winter.

• When will we see a VC just strap an EEG to founders heads to track brain activity in their pitch?

• What does “sovereign” or “resilient” infrastructure actually mean in 2026?

• In particular, how does GPS work, where it breaks (jamming/spoofing), real-world attack examples, economic cost of outages, and what actually reinforces it (alt-PNT, timing, atomic/quantum clocks, redundancy)?

• How do frontier algorithms get “born” outside the LLM monoculture?

• For semiconductor/algorithm plays, what benchmarks matter, what “best-in-class” means, and which milestones are important to understand?

• Where is robotics heading — and what’s investable versus sci-fi?

• What is the simplest first-principles chain that makes this a venture-scale outcome if one assumption holds?

• Are we in a bubble?

• Is it better to raise now or in a year from now?

• December

• November

As ever, if anyone is thinking about these topics please reach out. All thinking is my own and reflects personal views, not those of my employer.

[Credit, thanks Peter for the excellent inspo - copied your idea].