Insights
Writing from Eldris on AI and the decisions around its use. This is commentary: argument and experience, clearly labeled. Original studies and reports belong to Research and Reports.
Latest insightEnterprise AI
Apple silicon and unified memory made meaningful local model deployment practical inside a personal computer. That change enabled OOMU, and deserves a larger place in enterprise AI strategy.
Inside organizations
Commentary for the people who decide how AI is bought, governed, and put to work.
An AI system can generate work faster than an organization can responsibly approve it. Leaders who promise human oversight need to fund and test the capacity behind that promise.
A provider's attractive inference price can conceal an expensive dependency. Enterprise buyers should make the ability to change providers part of the original investment decision.
Foreign AI providers need a proposition that a U.S. buyer can approve, operate and defend. Technical superiority is one part of that proposition.
Employers who want employees to reveal and help realize AI productivity gains should make a credible bargain about what happens next.
Executives should be willing to expand a few valuable applications and close many others. A strategy that cannot stop an AI project cannot allocate capital seriously.
A successful demonstration can leave the hardest deployment questions unanswered. Leaders should demand a pilot that tests who will operate the system, absorb its costs and own its consequences.
In people's own lives
Commentary on what people let AI do for them, and what products owe them in return.
An assistant that presents itself as working for the customer should make commercial influence visible and let the customer define what a good purchase means.
Consumer AI companies should earn the right to act through permissions people can understand, limits the product enforces and remedies people can actually use.
What we publish
Short senior analyses of consequential AI decisions, written for directors, executives, and the advisors around them.
Analysis connecting AI policy to what it does inside a working organization, rather than restating what a proposal says.
Observations from building and running OOMU: local and cloud economics, workflow design, and human approval. Experience, not study results.
The monthly brief
Eldris circulates a private monthly decision brief to a small group of senior contacts: directors, general counsel, government relations leaders, investors, and advisors.
To receive it, tick the monthly brief box on the contact form. It is a separate choice from anything else you send us, and you can ask to stop at any time.
Looking for evidence?
Commentary is useful for thinking a question through. It is not evidence about how people adopt AI. For that, see the research programs, commissioned studies, and reports.
The writing is the general case. An advisory engagement is the specific one.
