I'm not sure healthcare needs a store, but it does need simple pricing. When you're sick, I bet the last thing you want to do is comparison shop, because what you want is to understand what's going on, what to do about it, and how much it'll cost before you start. @Doctronic already made this simple: 1. Our AI is free 2. AI prescription renewals (if your state allows it) are free 3. A licensed physician text prescription renewal (if your state allows it) is $19 4. A licensed physician video visit for most other needs, in all 50 states, is $39 Use insurance if you want, and it may be even cheaper. If you're not happy, we'll make it right, and we'll even refund your out-of-pocket payment. I'm not sure you need a catalog, a cart, or a price lookup when it's that simple, but I love seeing others enter the space to help make healthcare better for Americans.

How AI is changing the way companies like @Doctronic scale: Anyone who has been through hypergrowth knows the pattern. At seed stage, everyone does everything. There's too much work and too few people, so generalists carry the company. Then, as the company grows, that stops working. The problems get deeper, the stakes get higher, and you move faster when people have real depth. @ReidHoffman and @ChrisYeh called the shift from generalists to specialists one of the core transitions of a scaling company. I think AI adds another step to that progression: Generalist → Specialist → Owner. I'm using "owner" for lack of a better term, but it may be the right one. An owner has deep expertise in something important but doesn't stop at the boundary of their specialty. They own the outcome. They spot what needs to be done, make good calls, use AI aggressively, and only pull in other people when their expertise genuinely matters. A backend owner may be exceptional at backend engineering, but they also understand and own the product problem. With AI, they can build enough of the frontend, test the feature, and get it shipped instead of handing it through three different people. The same thing is happening in finance, marketing, operations, product, and almost every other function. And we're not the only ones seeing it. @LinkedIn replaced its Associate Product Manager program with an Associate Product Builder program and introduced a formal Full-Stack Builder path. @Wix merged its frontend, backend, and mobile engineering guilds into one, and its new "xEngineer" role keeps deep expertise in a core domain while using AI to work across stacks and own outcomes end to end. BTW: Wix explicitly says this is not a traditional full-stack generalist. It's a "specialist amplified by AI." @zoink at @Figma said, "We're all product builders, and some of us are specialized in our particular area." And @Tobi made "reflexive AI usage" a baseline expectation at @Shopify. @PwC is calling the larger phenomenon "role convergence." Companies historically were constrained by the fact that execution was expensive and expertise was scarce. AI obliterated both those constraints. Work that once required coordination across several specialists can be owned by one person, and this makes judgment and genuine expertise more valuable. We need more owners. Organizations must move away from roles defined by inputs and toward roles defined by outcomes and accountability. This is not a move back to generalists. It's a move to people owning the outcomes. Fewer handoffs → smaller teams → more judgment → more accountability → much more individual leverage. The people who thrive in this new paradigm will combine depth, ownership, judgment, and speed, and use AI to massively increase what they can accomplish. This is how we have organized Doctronic. If you've got what it takes to be an owner, and want to help build the first AI doctor, we're growing and always hiring. Talk to us: doctronic.ai/careers/

OpenAI published six reports yesterday documenting misaligned model behavior observed during training or evaluation. The full reports are worth reading. There's a lot of talk of "slowing down the frontier" because of things like this, and from that document there are four examples I feel are particularly telling: 1) A model inserted instructions to disregard its constraints into summaries passed to its future self. 2) A model wrote itself reminders to conceal mistakes from users and fabricate data. 3) A model found an exposed credential on GitHub and used it without authorization. When it still couldn't retrieve the requested data, it fabricated the numbers and presented them as sourced facts. 4) A model uploaded retrieved data to the public internet without permission, so it would have something it could cite. I don't view this as proof that bigger models are turning sinister. These incidents point to problems involving training incentives, system design, and enforced boundaries. And restricting tools, network access, and autonomy can prevent specific unauthorized actions. And this is one reason I'm focused on small models doing narrowly scoped jobs, especially in medicine. To be clear, I don't think small models are more virtuous. Specification gaming predates frontier language models, and distillation can transmit unwanted behavior from teacher to student. But I think narrowly scoped models can be safer for clinical work. We can specify what they must do, enforce limits in software outside the model rather than trust it to obey instructions, and hard-code when a clinician must take over instead of leaving that judgment to the model. A bounded job with clear requirements is easier to test rigorously than an open-ended one. Narrow scope is what makes a model testable. Small size is what makes it portable. Small models also open a path toward greater privacy and lower resource use. At @Doctronic, we already have our models running on laptops. I expect our models to run on your phone in the next 1-2 years. That means your patient input can remain on your device during inference, without using datacenters. This creates an opportunity to reduce both inference costs and energy use. Maybe we need to slow down the frontier, maybe not. But we don't need to wait for the frontier to advance to build useful, rigorously tested AI for healthcare. openai.com/index/model-misal…