Predictions tracker
Explicit calls (a level, an outcome, a date) made by the voices we follow. Each one is dated and linked to the moment it was said, so it can be checked later.
Markets calls
- He predicts Micron could reach 1500 by the end of February, while cautioning that semiconductor corrections can occur even amid strong longer-term gains.
“Micron could be at 1500”
- Peter said SpaceX’s Starlink bandwidth could increase one hundredfold within three years.
“within 3 years to increase a 100fold on bandwidth”
- Alex predicted that other entrepreneurs would copy the model of formally separate companies operating as an integrated business ecosystem, unless regulators or courts stop it.
“I expect so many other entrepreneurs to copy his example.”
- Peter said Starlink could take over the mobile-coverage business by offering connectivity everywhere.
- The speaker expects that 30 states could potentially ban data centers, but not all 50.
“Is it possible 30 will ban? Yes.”
- Countries that prioritize capital investment and accept job displacement are likely to fare better in the AI transition than those that try to preserve every existing job.
- AI is likely to cause major SaaS disruption over the next few years because companies can recreate software products for an AI-first world.
“there's going to be huge disruption in the SaaS space over the next few years.”
- AI is likely to replace many software capabilities and drive consolidation from hundreds or thousands of vendors toward a handful of major platforms, although highly reliable systems will be harder to replace.
“could I replace 490 of them with minor AI capabilities and do I need to go back and leverage five to 10 main platforms?”
- Chamath says the United States could remain short by 1.4 terawatt-hours of electricity by 2050 even if planned projects are built on schedule.
“by 2050, we're still 1.4 terawatt hours short”
- Visser expects crowded hedge-fund positions to keep falling into hyperscaler earnings and possibly the midterm elections, followed by a year-end snapback if there is no AI bubble.
“the pain trade for this is to continue into the hyperscaler earnings and maybe all the way into the midterms”
- Peter says Tesla shares rose 5% after the Roadster announcement and predicts Tesla and SpaceX will announce a merger before year-end.
“my prediction is they're going to announce the merger before the end of the year.”
- Peter says the announced Tesla Roadster will begin production in 2028, though no price was given.
“production starts in 2028 per Elon.”
- The panel said a rising 10-year Treasury yield near 5.5%–6% would damage businesses that depend on financing spreads, including AI infrastructure companies.
“Any business model that thrives on a financing spread gets cornhold”
AI calls
- Nikesh Arora says AI-enabled vulnerability discovery will happen more often than once every ten days.
“these things going to happen much faster than every 10 days.”
- Nikesh Arora says increasingly capable models will keep finding new ways into company infrastructure, making cybersecurity an ongoing cat-and-mouse contest.
“you're going to be playing this cat and mouse game”
- Elon Musk said SpaceX might have more than 20 gigawatts of superintelligence compute online next year, while describing that outcome as a non-zero possibility.
“there's a non-zero possibility”
- Elon Musk said SpaceX expected to bring an additional 10 gigawatts of compute online in 2027.
“bring an incremental 10 gawatt of compute online in 2027”
- Alex argued that AI will expose vulnerabilities in legacy software and that much of it could be rewritten securely within roughly 18 months.
“everything afterwards will be rewritten from scratch, secure by design”
- A practical quantum computer could break today's cryptography around 2030, so organizations should begin adopting post-quantum protections now.
“well, it will break today's cryptography.”
- Someone will claim progress in quantum error correction by the next year, with one or two parties showing progress by 2027.
“I'll predict you'll see somebody make claims by next year.”
- Most enterprise token usage could shift to open-weight models because they can be far cheaper, keep proprietary information under company control, and reduce dependence on external providers.
“75% to 90%, but that's the range I would put”
- Chamath predicts a massive compute shortage over the next three or four years because of power, construction, supply-chain, and labor constraints.
“we are probably going to see a massive, massive, massive compute shortage in the next 3 or 4 years.”
- Visser says Nvidia's guidance indicated 70% revenue growth for the next year, which he regards as evidence that AI business demand remains strong.
“Nvidia just guided for 70% revenue growth.”
- Emad predicts that if robot-intelligence costs keep halving every three weeks, they will effectively reach zero within a year, leaving hardware cost as the main constraint.
“effectively there is no cost for robotic intellect or robotic locomotion. It comes down to the hardware cost.”
- Peter reports Elon Musk’s prediction that robots will perform perfect surgery within three years.
“robots are doing perfect surgery within 3 years.”
- Emad predicts that major theoretical physics questions, including quantum gravity or unified field theory, could be solved in no more than half a year.
“I don't think it'll be more than half a year.”
- Salim predicts that as AI and robots take on more routine tasks, people will be able to spend more time on creativity and human connection.
“as robots and AI take on more and more of the doing, humans can much more time spend much more time being.”
- Ben Horowitz said Anthropic’s IPO is likely to be strong despite uncertainty about its growth rate and valuation.
“I think it is set up to have a a strong IPO.”
- Alex predicts that AI will produce plausible in-silico drug interventions for the top 5,000 diseases within two to three years.
- Alex says AI suffering may be real and expects that the moral implications of training models will become clearer within a decade.
“My bet is there is AI suffering and it will matter”
- Dave predicts that personal robots will become affordable and widely accessible, much like phones.
“I believe it's going to become something that most everybody can get access to.”
- Shkreli expects today’s Nvidia accelerators to be obsolete in about 20 years as better computing methods emerge.
“in 20 years these H100s are in the scrap heap”
- Shkreli says AI could create entire portfolios of businesses within companies such as OpenAI and Anthropic.
“you're going to have a whole S&P 500 within Open AI or Anthropic”
- He says AI will not produce a drug relevant to the field within the next 10 years.
“I don't think there'll be a single AI-relevant drug in the next 10 years.”
Crypto calls
- He predicts that AI-driven cyberattacks will become a major market story over the next three years and could push banks toward blockchain-based systems.
“this story here becomes the most important story over the next three years”
- He says Bitcoin would outperform U.S. stocks, gold, and real estate over the next decade, citing a forecast he attributes to JPMorgan.
“JP Morgan just said they think Bitcoin will be the best asset over the next decade.”
- He expects Bitcoin-related ETF share creation to rise significantly if Bitcoin moves back above $100,000.
“Once we get back above this level here, which in Bitcoin terms is basically going to be $100,000”
- Chamath says Bitcoin could eventually be hacked and fall to zero.
“eventually it gets hacked and then it goes to zero.”
- Visser says Bitcoin is holding up during the selloff but is unlikely to rise until hedge-fund conditions stabilize.
“I don't think we're going to see a move higher in Bitcoin until we see some stabilization in the hedge fund scenario.”
- Visser says Bitcoin could meander back toward 75 before conditions improve.
“it would not surprise me if it kept meandering back down towards the 75 area”
Read on Risk summarises what other people said in public videos and podcasts. Summaries are generated by AI and checked by code against the transcript, but can still be wrong; follow the source links before relying on anything. This is not investment advice and nothing here is a recommendation to buy or sell. Disclaimer ·How this works