The mythos of Mythos and Allbirds takes flight to the neocloud

Daniel Leitnack, CEO of Prediction Guard, and co-host Chris Benson discuss Allbirds' transition into an AI company, highlighting its significant growth from 2016 to 2021 and subsequent decline, leading to the sale of its footwear assets in March 2026. Despite exiting the shoe market, Allbirds retained its structure and cash reserves, raising speculation about its future in AI infrastructure. The surprising pivot resulted in a 700% surge in Allbirds' shares, suggesting that other struggling companies might consider similar strategies.

The conversation shifts to market trends, noting a growing acceptance of AI-focused business models. Allbirds is compared to Nike, prompting questions about Nike's potential branding as AI-centric. The term "neocloud" is introduced, referring to cloud infrastructure designed for AI workloads, which contrasts with traditional cloud services. Concerns arise regarding Allbirds' entry into the neocloud space, particularly about their AI expertise and the implications of their $50 million investment in a competitive market.

The discussion explores trends in AI deployment, emphasizing the centralization of compute resources in data centers and the rise of embedded AI. The potential of "far edge" computing is highlighted, with established players likely dominating the neocloud strategies, impacting profitability. The Mythos model from Anthropic is introduced as a next-generation frontier model effective in identifying security vulnerabilities in software packages.

The conversation touches on Project Glasswing, a closed security initiative involving around 40 companies utilizing mythos to secure their systems. There is ongoing uncertainty about the future of mythos, echoing past discussions about AI model releases. The increasing sophistication of threat actors in cybersecurity is contrasted with Anthropic's safety-oriented approach, emphasizing the necessity of governance and control in the AI landscape.

The term "token maxing" emerges, discussing how AI tools have transformed coding workflows. Meta's strategy gamifies developer engagement with cloud tools, leading to a competitive environment for token usage. Concerns arise about organizations not investing enough in AI and the need to establish boundaries in AI token usage. A recent federal court ruling highlights that AI-generated outputs are not protected by attorney-client privilege, complicating the legal landscape surrounding AI tools.

Concerns about AI chat logs in legal contexts are raised, noting that users may fear their conversations could be used against them in court. The discussion extends to the legal and medical fields, where courts may request access to chat logs, prompting caution in handling confidential information when using AI systems. The potential for AI chat systems that ensure no records are kept is suggested, inviting insights on the legalities surrounding privacy and record-keeping in AI chat systems. The evolving nature of confidentiality rules across communication technologies is acknowledged, presenting potential market opportunities.