Keep pulling the thread on Aaron Levy.
Aaron Levy believes it would be difficult for a standalone company to build a massive business solely by offering AI models to enterprises due to competitive pressure from hyperscalers like AWS, Microsoft Azure, and GCP.
In internal tests conducted by Box, Google's open-weights model Gemma performed only a few percentage points worse than the commercial Gemini model for data extraction from enterprise documents.
Aaron Levy believes the technology industry is overly focused on consumer AI applications, while the larger financial opportunity, worth hundreds of billions of dollars, lies in B2B AI.
Aaron Levy predicts AI will enable software categories to expand their Total Addressable Market (TAM) by capturing budget from adjacent services markets, citing the potential for contract management software to grow from a $2 billion to a $20 billion market by absorbing legal services spend.
Enterprise customers are adopting AI at a pace 5 to 7 years faster than they adopted cloud computing, bypassing the initial resistance and justification phase.
The vision of AI agents replacing all SaaS applications, as suggested by Klarna's experiments, is unlikely because most enterprises do not want the liability of running their own AI-generated critical systems.
AI model providers like Anthropic will not take legal or operational responsibility for applications built with their models, in contrast to SaaS vendors like Workday who are accountable for their products.
Aaron Levy predicts that for any major AI model breakthrough achieved by a closed-source provider, an open-source equivalent will be available within three months.
Aaron Levy considers Mistral to be the only "true model company" currently operating.
Aaron Levy does not categorize OpenAI or Anthropic as "model companies," but rather as "AI foundation, frontier lab companies" that have products, models, and services.
Aaron Levy identifies Google's strategy for its core search product in the age of AI as a critical strategic question for the company.
Aaron Levy views Apple's strategy for Siri, specifically the decision of whether to build its own AI technology versus partnering with other players, as a key strategic challenge.