A Stanford report on 51 enterprise AI deployments highlights what works — but also what’s missing. While human-in-the-loop models dominate successful use cases, the claim that “messy data isn’t a blocker” contradicts real-world experience. Notably, the report avoids the term “hallucinations” entirely, despite citing reliability issues in 27% of cases. Even among winners, AI still struggles with consistency.

Sift has raised $42M to build the data infrastructure layer for physical AI. As machines generate massive, unstructured sensor data, AI still struggles to interpret it. TheMarketAI Take: the challenge in robotics isn’t just better models but making the physical world legible. Bridging that gap may be key to scaling AI beyond software.

A new article from Thinking Machines Labs may hint at where the record-setting startup is heading. The piece explores why LLMs remain nondeterministic even with temperature set to zero tracing the issue to floating-point math across multiple cores. With reports that 95% of AI pilots fail, Thinking Machines seems to be signaling that unpredictability is more than a quirk it’s a core obstacle to scaling AI reliably.

We tested OpenAI’s new GPT Agent on a real task: calculating our AI SC 15 Index. The result? Fast, accurate, and low-friction. As tools like this improve, startups built on similar workflows may struggle to keep up — a trend we explored in our recent piece on AI’s rapid pace.

New Homebase survey shows 64% of SMBs are already using or testing AI, with growing familiarity around agentic tools. But concerns around privacy (61%) and errors (47%) remain key barriers. The findings highlight both the accelerating adoption of AI by small businesses and the persistent trust issues slowing deeper integration.

What happens when AI systems generate plausible-sounding but completely false information? Welcome to AI hallucinations—fabricated content that's not grounded in context or world knowledge. As AI becomes more sophisticated, distinguishing accurate responses from confident fiction becomes crucial for anyone using these tools.