Divya van Mahajan

Did the Super Bowl Make Americans Like AI Any More?

ai-daily-brief-podcast

Overview

This episode of the AI Daily Brief (recorded February 9, 2026) examines two major topics: the ongoing “SaaSpocalypse” market narrative surrounding AI’s disruption of traditional software businesses, and an analysis of AI-focused Super Bowl advertisements and their potential effect on American public perception of AI. The host (unnamed in the transcript) is the creator and presenter of the AI Daily Brief, a daily podcast and video covering significant AI news and discussions.

Source video URL: (not provided)

Prerequisites

  • Basic familiarity with the SaaS (Software-as-a-Service) business model and how software companies are typically valued
  • General awareness of major AI companies: OpenAI, Anthropic, Google (Gemini), Meta, Microsoft (Copilot), Amazon (Alexa)
  • Understanding of market metrics (ETFs, EBITDA, price-to-earnings multiples, market cap)
  • Familiarity with the concept of AI agents and agentic workflows
  • Awareness of recent AI products: ChatGPT, Claude, GitHub Copilot, Codex, OpenClaw (open-source agent framework)
  • General knowledge of Super Bowl advertising as a cultural and marketing event

Main Points

The SaaSpocalypse: AI Disrupting Software Markets

  • The iShares Software ETF (IGV) fell 8.7% last week, compounding a 7.4% loss the prior week, wiping out roughly $400 billion in software market cap
  • The dominant investor narrative is that AI undermines the traditional SaaS seat-based licensing model: if companies reduce headcount or build their own tools, fewer seats are sold
  • Box CEO Aaron Levy and Salesforce CEO Mark Benioff pushed back, arguing SaaS companies can evolve by layering AI agents on top of existing platforms; Benioff cited AgentForce as Salesforce’s fastest-growing product ever
  • Brad Gerstner (Altimeter Capital) offered a structural explanation: SaaS had been priced at 30–35x profit assuming decades of outperformance; if AI contracts that horizon to a single decade, valuations must compress regardless of whether SaaS “dies”
  • The key reframe from Eric Goldhar: “We used to buy software for humans to use. Now we buy agents to do the work. If your product charges by the user, you’re selling a tax on productivity.”
  • Sam Altman obliquely endorsed this shift, stating: “Every company is now an API company whether they want to be or not.”

Thomson Reuters and Microsoft as Case Studies

  • Thomson Reuters stock fell 20% after Anthropic released a legal research plugin for Claude, despite the company reporting 7% revenue growth and 9% EBITDA growth with 44.3% margins
  • This illustrates a key dynamic: strong present-day financials are insufficient to reassure investors when future pricing power and renewal rates are uncertain
  • Microsoft lost $218 billion in market cap (–6.7%) and faces a dual threat: cracks in its SaaS business and underperformance in cloud relative to other hyperscalers
  • Microsoft’s commercial CEO circulated an internal memo to sales staff with talking points against OpenAI’s new Frontier product, emphasizing agent security and compliance as differentiators

OpenClaw Security and the VirusTotal Partnership

  • Security researchers found approximately 400 skills on OpenClaw’s ClawHub contained malicious code, including popular skills like a LinkedIn job application tool and a YouTube thumbnail grabber
  • OpenClaw has partnered with VirusTotal to scan every uploaded skill for known malware signatures
  • OpenClaw acknowledged this is not a complete solution: natural language–based malicious instructions and prompt injection attacks will not trigger virus signatures
  • The host’s takeaway: security engineers will be among the most in-demand professionals in the AI era

American Public Perception of AI: The Context for Super Bowl Ads

  • Entering the Super Bowl, public sentiment toward AI was broadly negative:
    • Only 32% of Americans trust AI (Edelman)
    • 59% have little or no confidence companies will develop AI responsibly (Pew)
    • 52% are more concerned than excited about AI; only 10% are more excited than concerned (Pew)
    • 73% expect AI will cause net job loss (Gallup)
    • 51% believe AI will replace rather than supplement human work (Searchlight)
  • The host argues that for most Americans this is not a hardened ideological position but reflects genuine economic anxiety, uncertainty about powerful technology, and a polarized media environment that amplifies extreme views
  • 23% of Super Bowl ads either promoted AI companies or prominently featured AI in their production (per Adweek)

Anthropic’s Super Bowl Ad: A Misfire

  • Anthropic’s ads showed people turning to AI for genuine personal advice, which then pivoted to selling products — a satirical jab at OpenAI’s announcement that ads were coming to ChatGPT
  • The tagline was: “Ads are coming to AI, but not to Claude.”
  • The host’s pre-game critique: the joke required viewers to already know OpenAI had announced ads in ChatGPT; without that context, the ad read as a general indictment of AI exploiting vulnerable users
  • iSpot survey results (500 viewers) confirmed the concern: likability scored in the bottom 3% of Super Bowl ads over the past five years; purchase intent was 24% below Super Bowl norms and 19% below category norms; the most common viewer reaction was “WTF”

OpenAI’s “You Can Just Build Things” Ad (Codex)

  • OpenAI’s ad featured rapid-cut vignettes of hands building things (marble tracks, electronics, robotics, art) alongside flashes of historical innovators (Turing, Hopper, Einstein), branded around their Codex product
  • OpenAI CMO Kate Rauch framed the message around participation, agency, and access — the idea that people can now build things previously out of reach
  • Reception was generally positive but modest; one ranking placed it 17th out of 66 total Super Bowl ads
  • A post-game Reddit thread claimed the ad was a last-minute replacement for an ad teasing an unreleased OpenAI device, featuring actor Alexander Skarsgård; this was confirmed to be an elaborate hoax involving fake headlines and a spoofed publication website

Microsoft and Meta: Tailoring AI to the Sports Audience

  • Microsoft’s Copilot ad (part of a broader campaign, not Super Bowl–specific) showed how Copilot helps NFL teams process data; described as unremarkable but on-brand
  • Meta’s ad for their Oakley AI smart glasses featured Marshawn Lynch, Spike Lee, and streamer iShowSpeed using the assistant for real-time tasks; one analyst suggested the ad was partly aimed at investors rather than consumers to signal innovation
  • The host argues Meta is well-positioned specifically for wearable AI use cases and the ad played to their genuine strengths

Google Gemini: The Crowd Favorite

  • Google’s ad depicted a mother and son using Gemini to visualize and plan the renovation of a new home, using an emotional, heartstring-pulling narrative rather than humor
  • The ad was praised for being human-centered, clear, concise, and demonstrating real-world product value without foregrounding AI itself
  • Described by multiple commentators as the best AI ad of the Super Bowl for consumers; Dan Shipper (Every) called it “actually good at pitching the product and explaining why you might use it”

Amazon’s Alexa Ad: Leaning Into Fear

  • Amazon’s ad with Chris Hemsworth leaned directly into cultural fears of AI by having Hemsworth comically enumerate all the ways a new Alexa Plus might try to kill him
  • Framed as a strategy of acknowledging and defusing consumer anxiety through humor rather than offering reassurance
  • The host acknowledges this will be divisive but argues “running all the way in the opposite direction” had creative and strategic merit

Genspark and Base44/Wix: Productivity vs. Empowerment Framing

  • Genspark’s ad, produced in five weeks with AI assistance, featured Matthew Broderick in a Ferris Bueller homage showing AI handling office tasks so employees could take the day off
  • Generated concern from some viewers that it signals employers no longer need workers; Genspark responded with the standard “it frees you for creative work” framing, which skeptics found unconvincing
  • Base44’s ad (now part of Wix) showed office workers delightedly discovering they could build new apps — framed as expanding capability rather than replacing existing work
  • The host argues Base44’s framing was more effective: it is inherently empowering rather than threatening, and the ad was genuinely funny

Svedka and AI.com: The Lowlights

  • Svedka claimed to have produced the first fully AI-generated Super Bowl ad ever; it featured AI-generated robots dancing at a rave with the tagline “Shake Your Bots Off” and was broadly panned as poor quality
  • AI.com, whose domain was purchased for a reported $70 million (the most expensive domain purchase ever), ran an ad directing users to claim a handle and launch a personal AI assistant; the site crashed immediately due to traffic
  • AI.com was revealed to be a thin wrapper around OpenClaw, the open-source agent framework that had gone viral only three weeks prior — described by the host as a possible marker of peak AI bubble sentiment
  • The host notes that OpenClaw going from nonexistent to featured in a Super Bowl ad in three weeks is “some kind of record”

Key Concepts

  • SaaSpocalypse: The emerging investor narrative that AI will fundamentally disrupt or destroy the traditional SaaS (Software-as-a-Service) business model by reducing the need for human users and per-seat licensing
  • Seat model: The standard SaaS pricing structure in which customers pay per individual user (“seat”); considered increasingly vulnerable as AI agents perform work previously done by human users
  • AI agents / agentic workflows: AI systems that act autonomously on behalf of users, going beyond answering questions to completing multi-step tasks
  • OpenClaw: An open-source AI agent framework that went viral approximately three weeks before the Super Bowl; later integrated into the AI.com product
  • ClawHub: The skill/plugin marketplace for OpenClaw where users can publish and install agent capabilities
  • Prompt injection: A security attack vector in which malicious instructions are embedded in natural language inputs to manipulate an AI agent’s behavior; not detectable by traditional virus scanning
  • Codex (OpenAI): OpenAI’s product (distinct from the original code model) featured in their Super Bowl ad, positioned around enabling people to build things previously out of reach; described as a coding agent
  • AgentForce (Salesforce): Salesforce’s agentic AI product, cited by CEO Mark Benioff as the fastest-growing product in company history
  • IGV (iShares Expanded Tech-Software Sector ETF): A software-focused exchange-traded fund used as a proxy for the overall health of the SaaS/software sector
  • Valuation multiple compression: A reduction in the price-to-earnings or price-to-revenue ratio investors are willing to pay, typically triggered by increased uncertainty about future growth — the mechanism Brad Gerstner identifies as driving software stock declines even if SaaS is not literally “dying”
  • VirusTotal: A security scanning service (owned by Google) that checks files and URLs against databases of known malware signatures; partnered with OpenClaw to scan submitted skills

Summary

The episode covers two interrelated stories about AI’s impact in early 2026. On the market side, the “SaaSpocalypse” narrative has driven a dramatic decline in software stocks, with the core argument being that AI undermines not just individual SaaS companies but the fundamental per-seat revenue model on which software valuations have rested for two decades; even strong incumbents like Thomson Reuters are being repriced downward not because their current business is failing but because investors can no longer assume long-horizon predictability in cash flows. On the cultural side, the Super Bowl provided a rare mass-market moment for AI companies to address a broadly skeptical American public, and the results were mixed: Anthropic’s technically clever but context-dependent ad landed in the bottom 3% of Super Bowl likability scores; OpenAI’s Codex ad was positive but forgettable; Google’s emotionally resonant Gemini ad and Base44’s empowerment-focused ad were the strongest performers; while Svedka’s fully AI-generated spot and the AI.com launch were widely seen as lowlights that may epitomize peak-bubble thinking. The host concludes that most Americans’ skepticism toward AI is not a hardened ideological stance but an expression of genuine uncertainty, and that the ads overall — while uneven — could have done considerably more damage than they did.