Product Builders

The episode introduces two potentially useful axes (system capacity and output complexity) and references Microsoft's 19 guidelines for human-AI interaction, but these are sketched rather than unpacked. Most of the runtime is consumed by vague, circling discussion with little net new per minute.

“It is like those first two things that I said. It's like systems capacity, um, the enormous amount of users that we can think of. And it's also that the system is learning along the way.”

“We are super happy with using AI as not needing to intervene, just letting it do it work. But we are also sometimes happy to have it work on a very great scale with very dense and intense and detailed human, um, human reach, human oversight over it.”

The 'cognitive contracts' framing around naming (autopilot vs. assistant vs. copilot) is a genuinely useful lens, and the observation that practitioners resort to fairy-tale metaphors when conceptual vocabulary fails is a fresh aside. Otherwise the episode recycles familiar HCI and responsible-AI discourse.

“And names carry cognitive contracts. Some contracts with people and autopilot implies you can disengage as a person.”

“we go to this fairy tales about something magical or dreams...these imaginations of something material suddenly becoming alive and we forget about the whole variety of different levels of interaction that can happen in between.”

Agata Jawosinska holds real academic credentials in HCI and leads a cross-country responsible-AI index project, making her a credible researcher. However, she is a PhD-level academic rather than a practitioner who has shipped AI products at scale, which limits the practitioner-facing value for a B2B operator audience.

“Agata Jawosinska researches human data interaction and teaches human computer interaction at SWPS University in Poland.”

“She is regional project leader for the Global index on responsible AI, tracking progress in responsible AI, um, across 138 countries right now.”

The Tesla autopilot incident provides a concrete anchor and the Microsoft 19-guideline corpus is named, but these are the only substantive specifics in 24 minutes. The two-axis framework is described abstractly without named product examples, case data, or metrics that a product operator could act on.

“a Tesla on autopilot drove for a 7:37, I suppose, minutes without the driver touching the wheel before the crash”

“Microsoft actually develops 19 guidelines for human AI interaction”

The host sets up questions with real intellectual intent (naming as design failure, agentic AI shifting the framework) and occasionally surfaces productive tension, but largely accepts vague answers without drilling into specifics and lets the guest meander without redirection.

“do you think that names we give to AI systems are themselves a form of design failure? And what would it look like to name them more like? I don't know honestly. Or better.”

“I'm not fully convinced that I am not a little bit afraid.”