CHAPTERS
- 0:07 – 0:20
2025 forecast kickoff: rapid-fire predictions format
Garry opens a New Year’s Day “lightning” episode focused on quick, high-conviction predictions for 2025. He tees up the panel and hands off to Diana to start.
- •New Year’s Day episode framing
- •Goal: predictions for what happens in 2025
- •Lightning/short format setup
- 0:20 – 0:50
AI’s Nobel momentum—and the next prize to watch
Diana highlights how AI already influenced major scientific breakthroughs, pointing to recent Nobel recognition in physics and chemistry. She speculates that AI-driven work could plausibly claim additional top-tier awards next, with math or economics being the most likely targets.
- •AI-linked Nobel recognition in physics (deep learning) and chemistry (AlphaFold)
- •AI accelerating scientific discovery and “abundance”
- •Speculation: AI could win/drive another major prize
- •Math and economics seen as more plausible than peace or literature
- 0:50 – 1:15
Collaboration at the frontier: AI + elite mathematicians
Diana zooms in on the idea of AI contributing to math breakthroughs, citing high-profile collaboration efforts. The discussion frames 2025 as a year when research partnerships between leading AI labs and top mathematicians could yield major results.
- •Potential for AI to contribute to mathematical discoveries
- •OpenAI collaborations referenced with Terence Tao
- •Broader wave of researchers experimenting with AI-assisted proof and reasoning
- 1:15 – 1:45
Prediction: stablecoins go mainstream for everyday purchases
Harj makes a bold, concrete call: within a year, stablecoin payments become common enough that most people will have bought something with one. He argues stablecoins are the first truly compelling mainstream crypto use case beyond speculation.
- •Claim: everyday stablecoin purchases (coffee/book) become common
- •Stablecoins positioned as practical crypto utility vs. speculation
- •Time-bound prediction anchored to consumer behavior
- 1:45 – 2:46
Why payments are hard: stablecoins as a marketplace bootstrap
Harj explains that payment networks behave like two-sided marketplaces, which are notoriously difficult to start due to chicken-and-egg dynamics. He argues crypto already solved one side—consumer wallets—making merchant adoption the remaining, more tractable hurdle.
- •Payments as two-sided marketplaces (consumers + merchants)
- •Chicken-and-egg challenge parallels startup marketplace building
- •Hundreds of millions already have wallets—consumer side is “pre-bootstrapped”
- •Merchant acceptance becomes the key go-to-market task
- 2:46 – 2:58
Merchant adoption is the unlock: USDC acceptance and sales execution
The group narrows in on the practical path to mainstreaming: getting merchants to accept stablecoins like USDC. Harj argues merchant onboarding is simpler because there are fewer merchants than consumers and it resembles a traditional sales problem.
- •Merchants are fewer and easier to target than consumers
- •USDC cited as an example stablecoin
- •Adoption framed as a ground sales distribution challenge
- •Expectation of local retailers accepting stablecoin payments
- 2:58 – 3:40
Regulatory tailwinds and the case for stablecoins in finance
Harj points to improving regulatory sentiment as a catalyst for broader rollout. Garry adds that stablecoins could address weaknesses in today’s financial infrastructure, referencing a high-profile fintech failure as evidence for the need for better transparency and tracking.
- •Regulatory environment seen as more optimistic
- •Stablecoins framed as “desperately needed” in finance
- •Synapse failure cited: lost/unclear user funds tracking
- •Transparency/ledger concepts suggested as a solution direction
- 3:40 – 4:59
Garry’s macro-crypto bet: DOGE government cuts, rates, and crypto prices
Garry offers a circuitous but specific prediction linking government spending restraint efforts (“Doge” initiative) to interest rates and crypto valuations. He argues that if deficit spending falls, rates could decline, boosting risk assets and crypto—potentially including Dogecoin.
- •Prediction ties government spending cuts to interest rates
- •Lower deficit spending → lower rates (hypothesis)
- •Crypto prices described as correlated with interest rates
- •Dogecoin singled out as a beneficiary if the narrative holds
- 4:59 – 5:19
If austerity fails: deficit spending, higher rates, and the downside scenario
Garry outlines the counterfactual: if the spending-reduction effort fails, markets may face higher spending, potentially keeping rates elevated. He frames the outcome as pivotal because rates have dominated macro conditions in recent years.
- •Failure scenario: spending rises again
- •Higher spending could sustain higher interest rates
- •Rates “rule everything” framing based on recent years
- •Acknowledges uncertainty; expresses hope the plan works
- 5:19 – 5:42
Deflationary tech as the wild card: AI enabling lower rates without inflation
Harj adds a macro overlay: even if rates fall, inflation might not re-accelerate if technology—especially AI—drives strong deflationary pressure. He suggests AI could be the strongest deflationary force of the era, altering how policymakers and markets think about rate cuts.
- •Concern: lower rates can reignite inflation
- •Tech is broadly deflationary; AI potentially the biggest force
- •Possibility: rate cuts without major inflation resurgence
- •Connects AI progress to macroeconomic outcomes
- 5:42 – 6:28
AI prediction: real-time Zoom video calls with lifelike avatars
Garry predicts a leap from natural-sounding phone calls with AI to full video “Zoom calls” with an AI avatar in real time. The focus is on latency, lip-sync, and realism improving enough to avoid uncanny valley and feel like interacting with a virtual person.
- •2024: AI phone calls became natural due to lower latency
- •2025: video calls with AI avatars become viable
- •Key technical hurdles: latency, lip-sync, realism
- •Goal: interaction feels real rather than uncanny
- 6:28 – 7:02
A “3D Turing test” and closing the lightning round
The group jokes about a future where AI could host the show, framing the idea as a “3D Turing test.” Garry closes with New Year wishes and a sign-off, reiterating the episode’s quick-hit predictions format.
- •“3D Turing test” concept for convincing AI avatars
- •Humor about AI-generated episodes
- •Wrap-up: lightning predictions delivered
- •Happy New Year closing and farewell
