Most important take away

The AI bubble is reaching peak absurdity (Allbirds pivoting to AI as “Newbird AI”), while real infrastructure money flows to players like Fluidstack (18B valuation, 50B Anthropic contract) and Wave (self-driving). OpenAI appears to be making defensive acqui-hires (Hero, TBPN) as Anthropic dominates the enterprise/coding space, and a growing “AI anxiety gap” separates industry insiders from an increasingly suspicious public.

Chapter Summaries

The Allbirds “Newbird AI” Pivot

Allbirds, after selling its shoe business for $39M, rebranded as “Newbird AI” and is trying to become a computing infrastructure company. The hosts see this as the most emblematic (if not most important) story of 2026 — echoing the 2017 Long Island Iced Tea blockchain pivot and 2021 meme-stock craze. Stock spiked on the announcement; hosts predict even more ridiculous AI-fairy-dust pivots by summer.

Wave’s $60M Extension from Chip Makers

UK-based self-driving startup Wave raised a $60M extension from AMD, Arm, and Qualcomm on top of its earlier $1B+ Series D (Nvidia, Microsoft, Uber, SoftBank). Wave’s contrarian, hardware-agnostic, neural-network-based approach lets it work with any sensor/chip combo. Uber has committed ~$10B across autonomous vehicle partnerships, including a $300M milestone-based investment in Wave for a London robotaxi service. Nissan is a confirmed customer; Stellantis and Mercedes are investors.

Fluidstack’s $1B Round at $18B Valuation

AI data center startup Fluidstack is in talks for a $1B round at an $18B valuation, and reportedly has a $50B contract with Anthropic. Hosts note skepticism about whether all announced AI infrastructure projects will actually get built — Anthony says the real number will likely land “between the numbers announced and zero.” Example parallel: Lucid raised $4B via SPAC and still struggles.

OpenAI’s Shopping Spree (Hero & TBPN)

OpenAI acquired Hero (a personal finance startup, clearly an acqui-hire as the product is shutting down) and TBPN (a business talk show / new media company). Hero represents OpenAI’s search for new consumer products beyond ChatGPT; TBPN represents an image/PR play as OpenAI faces scrutiny (including a Ronan Farrow New Yorker piece on Sam Altman). Both signal OpenAI is anxious about sustainability and Anthropic’s enterprise lead.

Anthropic’s Rise and “Mythos”

Anthropic is winning in enterprise and coding (Claude Code dominated conversation at the HumanX conference). Anthropic announced “Mythos,” a model so powerful they claim they can’t release it publicly — Jerome Powell reportedly convened big banks to evaluate it as a cybersecurity tool. Despite Anthropic’s tension with the Pentagon, other parts of the Trump administration are engaging.

The AI Anxiety Gap

A Stanford report highlights a growing disconnect between AI insiders and everyone else. The public is increasingly suspicious while industry insiders (especially programmers) live in a transformed world. The attack on Sam Altman’s house this month underscores escalating tensions. Hosts call for taking criticism seriously without condoning violence.

Tokenmaxxing

New term describing the pressure at tech companies to maximize token/compute usage as a performance metric. Meta had a leaderboard (since shut down after leak) tracking who used the most tokens, which incentivized gaming (padding queries). Parasail raised $32M to help companies scale AI inference more cheaply. Uber’s CTO said they’ve already blown through their AI spending budget. Parasail generates 500 billion tokens per day.

Summary

Actionable Insights

On investment/market signals:

• Be skeptical of “AI pivot” stocks. Allbirds → Newbird AI is a textbook sign of late-cycle mania. Historical pattern matches Long Island Iced Tea → Long Blockchain (2017, later delisted) and the NFT/Web3 pump-and-dumps. Hosts expect more absurd AI pivots through summer 2026; treat these as speculative sentiment indicators, not real businesses.
• “Billion dollars” is small in AI infrastructure. The hosts repeatedly note that in AI/data center context, $1B is not a lot. Apply skepticism when evaluating deal sizes — look at whether a company can get GPUs, relationships, and scale, not just the headline number.
• Announced AI infrastructure buildouts will underdeliver. Expect actual construction to land between announced numbers and zero. Watch for quiet walk-backs from hyperscalers.
• Watch for S-1 filings later in 2026. These will reveal whether AI companies’ pricing is enough to build sustainable businesses versus being subsidized by investor capital. Sean flags this as the real question to watch.

Stocks/companies mentioned:

• Allbirds (BIRD) / “Newbird AI” — speculative pump on AI pivot; hosts clearly view this as not a real business. Treat as cautionary, not actionable long.
• Wave — private UK startup; notable as one of Europe’s more valuable startups. Serious Uber partnership with milestone-based $300M commitment for London robotaxi. Nissan confirmed customer.
• Uber (UBER) — committed ~$10B across autonomous vehicle bets. Long-game strategy. Uber’s CTO admitted blowing through AI spending budget — worth watching for margin impact.
• Nvidia, Microsoft, SoftBank — investors in Wave’s Series D.
• AMD, Arm, Qualcomm — invested $60M real equity (not in-kind) in Wave; signals chip makers want in on vehicle AI stack.
• Fluidstack — private; $1B raise at $18B valuation, $50B Anthropic contract. Emerging pattern: AI startups becoming dedicated infrastructure plays for frontier labs instead of competing.
• Lucid (LCID) — cited as cautionary tale: raised $4B, still struggling to scale.
• OpenAI — acquired Hero and TBPN. Seen as defensively reacting to Anthropic’s enterprise dominance.
• Anthropic — dominant in enterprise/coding; Claude Code is the tool people at HumanX were talking about. Announced “Mythos” model with government interest.
• Parasail — $32M raise; represents the emerging “token-efficiency” category. Worth watching as a wedge into AI cost optimization.

Career advice embedded in the episode:

• If you code, your competitive landscape has already shifted. Hosts note programmers’ daily work has been radically transformed vs. 2 years ago; most other professions have not. Implication: software engineers need to be fluent in AI coding tools (Claude Code being the current reference point) or risk falling behind. If you’re in enterprise/coding-adjacent roles, familiarity with Claude Code in particular is where the professional conversation is happening.
• “Tokenmaxxing” is now a workplace metric. At some AI-forward employers, visible AI usage is being tracked (Meta’s leaderboard). Career implication: if your employer tracks tokens, understand the metric — but also recognize it can lead to absurd gaming behavior. Real productivity, not token count, is the underlying signal.
• Domain expertise + AI infrastructure = premium positioning. Fluidstack-style “infrastructure-for-labs” plays and Parasail-style efficiency plays show startup opportunity in enabling rather than competing with frontier labs. If you can’t beat them, join them — applies to both startups and individual career bets.
• Be prepared for AI anxiety in customer/user bases. Stanford’s research confirms a widening insider/outsider gap. If you’re building consumer AI products, plan for suspicion and backlash, not just enthusiasm.

Category-level takeaways:

• Defense/government AI spending is wide open; both OpenAI and Anthropic are aggressively pursuing DoD contracts. Career opportunity in AI + defense.
• Acqui-hires in AI (Hero-style) are happening fast for anyone with serial-consumer-app or specialized skill sets. If you’re a founder of a modestly-performing AI-adjacent consumer startup, there is a buyer pool.
• Enterprise and coding are where the real money is; consumer chatbot revenue may not be enough to sustain OpenAI-scale companies.