Position: No (NaN contracts)Entry: 0% ($99.54)Current: 60% ($NaN)Max payout: $NaNP&L: $0.0000
Predictions
A public ledger of my predictions. If you're not willing to be wrong in public, you probably haven't thought hard enough.
Kalshi Positions
Live positions from Kalshi. Updated every 15 minutes.
Position: No (NaN contracts)Entry: 0% ($99.62)Current: 50% ($NaN)Max payout: $NaNP&L: $0.0000
Position: No (NaN contracts)Entry: 0% ($49.92)Current: 4% ($NaN)Max payout: $NaNP&L: $0.0000
Position: No (NaN contracts)Entry: 0% ($50.00)Current: 78% ($NaN)Max payout: $NaNP&L: $0.0000
Position: No (NaN contracts)Entry: 0% ($249.33)Current: 4% ($NaN)Max payout: $NaNP&L: $0.0000
Position: No (NaN contracts)Entry: 0% ($49.74)Current: 15% ($NaN)Max payout: $NaNP&L: $0.0000
Position: No (NaN contracts)Entry: 0% ($14.88)Current: 66% ($NaN)Max payout: $NaNP&L: $0.0000
Position: No (NaN contracts)Entry: 0% ($4.96)Current: 97% ($NaN)Max payout: $NaNP&L: $0.0000
Position: No (NaN contracts)Entry: 0% ($127.40)Current: 90% ($NaN)Max payout: $NaNP&L: $0.0000
Who will win MVP?↗
activePosition: No (NaN contracts)Entry: 0% ($29.55)Current: 92% ($NaN)Max payout: $NaNP&L: $0.0000
Position: No (NaN contracts)Entry: 0% ($50.00)Current: 31% ($NaN)Max payout: $NaNP&L: $0.0000
Long-Term Views
At least 3 of the top 10 US companies by market cap will be primarily AI companies by 2030
openConfidence: 75%Made: Feb 15, 2026Resolves: Dec 31, 2030
The rate of enterprise AI adoption combined with the infrastructure build-out suggests AI-native companies will capture enormous value. NVIDIA is already there, and companies like OpenAI, Anthropic, or their successors could join. The main risk is that incumbents (Google, Microsoft) absorb enough AI value to prevent new entrants from reaching the top 10.
Most knowledge work will involve daily AI collaboration by end of 2027
openConfidence: 85%Made: Feb 15, 2026Resolves: Dec 31, 2027
The trajectory of tool integration (Copilot in Office, Claude in IDEs, AI in email/search) makes this feel almost inevitable. The question isn't whether AI will be embedded in knowledge work tools, but how deeply. 85% confidence because adoption curves in enterprises can be slower than expected.
At least one AI lab will demonstrate a system that can reliably perform multi-hour autonomous research tasks by end of 2026
openConfidence: 70%Made: Feb 15, 2026Resolves: Dec 31, 2026
Agent capabilities are improving rapidly, and the research use case is particularly well-suited because it involves primarily text-based tasks with clear success criteria. The 30% doubt comes from reliability — 'reliably' is doing a lot of work in this prediction.