Nyeste episode
Bayes and Base Rates AI Forecasts and Probabilistic Thinking Mauboussin and Callahan 2026 This extended audio podcast is based on the Morgan Stanley Consilient Observer report “Bayes and Base Rates: How History Can Guide Our Assessment of the Future,” written by Michael J. Mauboussin and Dan Callahan and published on February 10, 2026.
T
Tid: 15:10
This extended audio podcast is based on the Morgan Stanley Consilient Observer report “Bayes and Base Rates: How History Can Guide Our Assessment of the Future,” written by Michael J. Mauboussin and Dan Callahan and published on February 10, 2026.
The report explores how investors can evaluate aggressive financial projections in the artificial intelligence sector using Bayesian reasoning and historical base rate data. Instead of relying solely on company specific narratives, the authors suggest starting with the statistical outcomes of similar companies over long periods of time and then updating those assumptions as new evidence emerges.
Using a dataset covering more than 75 years of U.S. public company history, the research shows that extremely rapid revenue growth at large scale is historically rare. As a result, some of the revenue expectations currently discussed for AI related businesses such as OpenAI or Oracle Cloud imply outcomes that have very low historical probability.
The report also examines the risks associated with massive infrastructure investment supporting the AI boom. Historical evidence from large project databases shows that fewer than ten percent of major infrastructure projects are completed both on time and within budget, highlighting the uncertainty surrounding the global buildout of AI data centers and computing capacity.
In addition to financial forecasts and infrastructure risks, the analysis considers strategic motivations behind large AI spending announcements. Firms may commit to large capacity expansions as a preemptive competitive strategy designed to discourage potential entrants and strengthen their leadership position in emerging markets.
Ultimately, the report argues that investors should adopt a probabilistic mindset when evaluating transformative technologies. By combining historical base rates with Bayesian updating, analysts can better assess the uncertain but potentially transformative impact of the generative AI revolution.
📈 Topics Covered• Bayesian analysis and probabilistic forecasting• Why base rates matter when evaluating AI revenue projections• Historical growth outcomes of U.S. public companies• Infrastructure risks in large scale AI investment projects• Strategic motives behind massive AI spending• Applying probabilistic thinking to emerging technology markets
Source reportMichael J. Mauboussin and Dan CallahanBayes and Base Rates How History Can Guide Our Assessment of the FutureMorgan Stanley Consilient ObserverPublication date February 10 2026https://www.morganstanley.com/content/dam/im/assets/publication/thought-leadership/consilient-observer/article_bayesandbaserates_ltr.pdf?1773434665395
Explore all ValueLetters playlistshttps://www.youtube.com/@ValueLetters/playlists
🔔 Subscribe to ValueLettersListen to extended audio podcasts based on leading investment research probabilistic thinking and long term analysis across technology markets and capital allocation.
#ArtificialIntelligence #GenerativeAI #BayesianAnalysis #BaseRates #MorganStanley #Mauboussin #TechnologyInvesting #AIInfrastructure #InvestmentResearch #ValueLetters
GUID: 037682e4-5172-43e1-aabe-dfd723a0a303
Utgitt: 13.3.2026, 21.47.35
