Henry Blodget on the Software Selloff Hysteria and the Problem for OpenAI
Most important take away
The current market panic around AI destroying software companies is hysteria — a repeat of the irrational sentiment swings that defined the 1990s internet boom. Henry Blodget argues that established software companies face disruption, not extinction, because enterprises require accountability and support that AI tools cannot currently provide, while the real unaddressed risk is not job displacement but the emergence of autonomous agentic AI systems operating with insufficient oversight.
Summary
The 180-degree narrative flip:
In May 2025, investors questioned how AI company valuations could possibly be justified. By March 2026, the concern has reversed completely: AI companies are now so powerful they’ll destroy every other company in their wake. Software and payment company stocks have been hit hard since January 2026. Blodget’s framework: extreme sentiment swings during major technology transitions are normal and often irrational. The same oscillation characterized the dot-com era, and disciplined long-term thinking outperforms reactive sentiment chasing.
OpenAI’s valuation problem:
OpenAI has risen from a $300 billion to an $800 billion valuation. Blodget is skeptical. The underlying bull thesis — that OpenAI will become the Google of the AI era — is, in his view, misguided. Google dominated search because search had natural monopoly characteristics and switching costs. AI assistants face a more competitive, commoditized landscape where Anthropic, Google, and others all offer credible alternatives. Blodget does not dispute that some AI companies could be worth enormous multiples of current value if the most aggressive scenarios play out, but he disputes the assumption that OpenAI specifically will capture dominant share. This creates both valuation risk for OpenAI investors and relative opportunity in competing platforms.
Why software companies will not disappear:
The apocalypse narrative fails for structural reasons. Enterprise software companies benefit from accountability and support structures that current AI tools cannot match. A business running thousands of employees on a critical platform cannot simply replace it with a junior developer and an AI chatbot — they need vendors who provide accountability, compliance, uptime guarantees, and support. Established software companies have the relationships, integrations, and institutional trust that take years to build. They face real disruption — pricing pressure, commoditization of basic features, reduced headcount for certain work — but not extinction.
Actionable investment insight: Software company selloffs may be overdone. The companies with mature enterprise relationships and irreplaceable integrations are likely to adapt and survive. Some current valuations may represent long-term opportunity for patient investors with conviction.
The real risk the market is ignoring — agentic AI:
Blodget shifts from economic disruption to safety. The emergence of agentic AI systems — autonomous agents operating in the real world with minimal human oversight — represents a more serious and underappreciated risk than job displacement. Markets are focused on the wrong problem. Economic disruption from AI is real but manageable and historically analogous to prior technology transitions. Autonomous agents making consequential decisions without accountability structures are a newer and less understood risk.
The core unknown driving valuations:
Whether AI disruption will actually reduce corporate profits remains genuinely unresolved. Markets have answered “yes” through lower valuations and higher discount rates across software. But this assumption is not proven. Companies may adapt, productivity gains may offset disruption, and enterprises may prove stickier than current narratives suggest. The increased discount rate reflects uncertainty, not certainty about negative outcomes. Investors who can tolerate that uncertainty and hold quality software names may be well positioned.
Market behavior during transitions:
The whipsaw from Monday’s broad risk-off selloff to recovery sentiment within days illustrates that sentiment is driving valuations more than fundamentals during this transitional period. Risk-off moves driven by narrative panic rather than fundamental deterioration historically create buying opportunities for long-term investors with conviction.
Wall Street’s repeated pattern:
Blodget (as a veteran of the dot-com era as an analyst, not just an observer) is explicit: Wall Street makes the same mistakes in every major technology transition — oscillating between extreme enthusiasm and extreme pessimism without doing the hard work of modeling outcomes. The current hysteria around software destruction is the bear-market mirror image of the 2024 AI euphoria. Both deserve skepticism.
Chapter Summaries
Chapter 1: The 180-Degree Flip — Framing the Conversation
Hosts open by noting the dramatic reversal in market narrative. Nine months ago investors questioned AI company valuations; now they fear AI will destroy all other companies. Software and payment companies are in selloff territory. Blodget sets up a measured perspective: he has updated his thinking on AI’s impact but maintains that extreme sentiment swings in both directions during tech transitions tend to overshoot.
Chapter 2: The Internet 1990s as the Correct Frame
Blodget draws the critical historical parallel: during the internet boom, a range of predictions from “it’s a fancy word processor” to “it will eliminate all jobs” coexisted simultaneously — and many of both turned out to be simultaneously true and false depending on the specific claim and timeframe. This framework helps contextualize the current volatility: don’t dismiss the disruption, but don’t extrapolate the most extreme version of it either.
Chapter 3: OpenAI at $800 Billion — A Skeptical Take
The discussion turns to OpenAI’s ballooning valuation. Blodget questions the “OpenAI becomes the Google of AI” thesis as the underlying justification. Google’s dominance rested on natural monopoly dynamics in search; AI assistants face a more competitive landscape. He notes that under the most aggressive AI dominance scenarios other companies could theoretically command enormous valuations too — but his skepticism is about whether OpenAI specifically will capture that value, and whether $800 billion is a price that makes sense absent that dominance.
Chapter 4: Why Software Companies Survive — The Accountability Argument
Blodget directly challenges the software apocalypse narrative. His core argument is structural: enterprises cannot replace years of accumulated integrations, relationships, compliance frameworks, and support with AI tools managed by junior developers. The disruption is real — pricing pressure, commoditized basic features — but the institutional trust and accountability moat protects established software providers from extinction.
Chapter 5: Enterprise Adoption Reality — Stickiness and Structure
The hosts probe how quickly enterprises actually adopt new tools. Blodget emphasizes the inertia built into large organizations: procurement processes, security reviews, compliance requirements, and risk aversion around mission-critical systems mean adoption curves are far slower than consumer technology. This structural stickiness is the floor under software company valuations even as AI accelerates disruption at the margin.
Chapter 6: The Real Risk — Agentic AI Safety
Blodget pivots to identify what he views as the more serious and underappreciated concern: not economic disruption from AI replacing jobs, but the emergence of autonomous agentic AI systems making consequential real-world decisions without sufficient oversight. Markets are pricing job displacement risk and ignoring governance/safety risk. He doesn’t claim this risk is imminent or inevitable but argues it deserves more attention than it’s receiving.
Chapter 7: Sentiment, Discount Rates, and the Core Unknown
The fundamental question driving valuations is whether AI disruption will actually reduce corporate profits — and this is genuinely unknown. Markets have increased discount rates across software as if the answer is yes. But adaptation is possible, productivity gains are real, and enterprises are stickier than current narratives suggest. The whipsaw from Monday’s selloff to early recovery illustrates that narrative panic, not fundamental deterioration, is driving current price action. This creates opportunity for investors who can hold conviction through the noise.
Chapter 8: Wall Street’s Repeated Blindness
Blodget closes with structural criticism: Wall Street consistently fails to navigate major technology transitions because it oscillates between extreme positions rather than doing careful scenario modeling. The current software selloff hysteria is the bear-market mirror of 2024 AI euphoria. Both represent sentiment extremes that should be discounted by investors applying historical perspective and long-term thinking.