The 2026 Prediction Bingo Update: Half Right. Fully Accountable.

Six months ago, we made a set of predictions about AI, markets, and startups in 2026. Now it's time to see how they're holding up.

Some aged surprisingly well: Google's AI search is becoming the default experience, AI specialists are going mainstream, and Canva continues its march toward IPO. Others... not so much.

Along the way, we unpack the biggest surprises of the year so far: Anthropic's stunning rise, growing skepticism around vibe coding, why AI wrapper startups are disappearing, and whether investors are underestimating how quickly foundation model companies can move into every category.

Then we’re coming in with a set of new predictions: physical AI, a Chinese breakout, video models, nuclear energy, and an AI-native creative studio that might just build the next MGM.

So who won our bingo card check-in? The answer depends on who you ask.

The views expressed on 60/40 are our own and are for informational purposes only. Nothing here constitutes financial, investment, or legal advice. Do your own research.

Timestamps

01:22 — The Bingo Card Is Open — Maria sets the format: six-month accountability round across AI, markets, and startups.

02:27 — Gemini Is the Default (In Most Places) — Aleksey's call on Google AI mode; one market where it's literally 100% right.

06:27 — The AI Wrapper Obituary — Prashanth on why wrapper companies are gone, and where the surviving moat actually lives.

09:37 — The Two-Year Runway Problem — Maria on how even regulated, data-heavy startups now have a compression clock on them.

14:35 — Vibe Coding's Accountability Crisis — Prashanth on what he's actually seeing in engineering teams and why the 90%-done first date gets very expensive.

17:08 — Anthropic's Surprise Dominance — How the company that stood up to the Pentagon became the product people stuck with.

22:34 — The IPO Pricing Clock — Aleksey's specific prediction: Anthropic raises real prices within three to six months of IPO.

31:14 — The Most Dangerous Startup Category — Wrapping up the retrospective with a fresh scan: what's most dangerous to build right now, and why 2018–2022 vintage startups face the hardest road.

36:50 — H2 Dark Horses: Video Models and Nuclear — Peter bets on large video models; Aleksey goes long on nuclear energy infrastructure as the only serious compute scaling answer.

Give us feedback or nominate topics & speakers at 6040podcast@techinasia.com.