Many models, some open: whose future is this survey describing?
October 4, 2026 · OSASI Commentary
The show's message is that the businesses getting real returns from AI now run a management layer over it: several models, rules for which one does what, a sovereignty strategy, and an eye on what it all costs. Average planned AI investment is up again, to US$210 million. The host tells solo operators why that matters to them too, but never puts a number on it. One detail helps: no organization in the survey earns less than US$50 million a year.
How a brief is made
A member reads an article or listens to a podcast and sends the snippets, key thoughts or concerns tied to a passage or a timestamp. An AI assistant, chosen for the task and the setup (Claude, ChatGPT, Codex, Gemini or another), searches the full article or transcript, pins the times, checks the quotations, and assembles a draft. The draft goes through rounds of review, and nothing is published until a person approves that exact version. The cues on each post start the player at the passage discussed. Commentary is OSASI's perspective, offered for discussion, not a summary of the source. Each brief is co-authored by a person and an AI assistant. AI assistant for this brief: Claude (Anthropic).
Source podcast · The AI Daily Brief
This episode prompted the commentary below. It is one perspective, offered for discussion.
The AI Daily Brief · October 2, 2026
What the Best Business AI Users Are Doing Different
Listening highlights · Nathaniel Whittemore, host, on his own. Times from Spotify's transcript and chapters.
Tap a time above to start the player there. If Spotify shows a card about the show first, close it and the player carries on at that time. Times are approximate transcript times; ads and episode editions may shift them. This plays the full episode, not an extracted clip.
Open episode in Spotify · Episode notes · KPMG: Global AI Pulse Q3 2026 (methodology on the last pages) · The Register: OpenAI apes AWS Marketplace while Anthropic remains stubbornly Microsoftian · Ara Kharazian (Ramp) on X: AI spend falls, but not because of open models · Baseten pricing: pay-as-you-go Model APIs, no monthly fee · OpenRouter FAQ: prepaid credits and provider data logging
Who is talking
Nathaniel Whittemore hosts The AI Daily Brief. This episode, published October 2, reads KPMG's Q3 Global AI Pulse; KPMG is also one of the show's sponsors. Quotations come from Spotify's transcript, and the survey figures from KPMG's report itself.
The show's own message
The businesses using AI best, Whittemore opens, "are building model routers, building organizational and data sovereignty strategies, and generally making their AI management layer much more robust", and in doing so "they are building a template that other businesses can follow."
KPMG now splits its sample into organizations still experimenting and those reporting established returns, and he reads the gap "like a map of where business AI users will go in general over the next several months." Among the established group, 86 percent run a formal layer of controls over their AI, 53 percent have an enterprise-wide sovereignty strategy (against 8 percent of experimenters), and 77 percent watch AI costs on a dashboard. Twenty-three percent have built some model routing. And average planned AI investment for the next 12 months rose from US$186 million in the first quarter to US$210 million, which he takes as a sign "we are still barely scratching the surface of what we will ultimately spend on intelligence."
Many models, some open
The trend runs toward several models rather than one, with open weights among them. OpenAI's new marketplace, which the last brief on open weights followed, gives enterprises "that multi-model flexibility that they are increasingly looking for and the ability to take advantage of cheaper open weight models", while OpenAI keeps "customer lock-in". The show adds a caution from Ramp's data: open models "remain less than 5% of business spend." Business AI is getting cheaper because OpenAI and Anthropic are competing, not because companies have moved to open models.
Who was asked
Whittemore does speak to "those who aren't in a big enterprise, for those solopreneurs among you or freelancers", with two reasons to pay attention: many of them serve these companies, and enterprise adoption "is going to have a dramatic impact on many parts of how AI evolves" for everyone else. Both are fair. Neither comes with a number.
KPMG's methodology fills in part of the picture. It surveyed 2,131 senior leaders, and every organization they represent earns at least US$50 million a year; in Canada, the United States and eight other large markets the floor is US$100 million, and the US tracking sample starts at US$1 billion. The US$210 million average is a big-company number. It says nothing directly about a two-person appraisal office.
The path for small businesses and solopreneurs
OpenAI's marketplace is open to eligible enterprise customers, spending part of an existing OpenAI commitment. A sole practitioner has no such commitment to spend. There are still three ways in.
Rent by the token, with no contract. Baseten, the partner behind OpenAI's open-model offer, also sells open models directly on a pay-as-you-go tier with no monthly fee, and services such as OpenRouter sell access to many open models from prepaid credit. The model is open; the servers and the data handling are the host's, so read the retention terms before a client file goes anywhere near it.
Run it on your own hardware. This is the other door: the model runs on a machine you control, and nothing leaves the office.
Share the capacity. Between renting from a stranger and buying alone sits pooling: owned hardware and tested methods, shared among peers. That is what the sandbox open-weights AI EOI and the OSASI Marketplace are for.
What this means for OSASI
What carries down to a sole practitioner is the shape, not the scale: more than one model, a rule for which one sees client files, knowing where the data goes, and knowing what it costs. None of that needs US$210 million. It needs a few hours, one of the three paths above, and someone to compare notes with.