Browse Comments — Raw (as collected)

Interesting demonstration. This highlights why autonomous medical triage is a regulated, high-risk application. Variability in recommendations from the same presentation can have real clinical and economic consequences. As a neuro-ophthalmologist, I would also question whether the observed differences truly reflect distinct clinical norms. The vignette lacks information that would typically be needed for disposition, making it difficult to know whether the recommendation is being driven by the clinical presentation itself or by assumptions associated with language and geography. In that sense, the finding may be even more important: the model appears willing to make a disposition recommendation despite substantial clinical uncertainty.

They evaluate how easily they can move from idea to outcome. When work requires constant context switching between tools, tabs, and platforms, the friction accumulates. The products that win are often the ones that simplify the entire experience, not just the individual task.

AI sin't just raising the bar. It's amplifying the gap between average users and expert users. The student who built and shipped an AI-powered tool while their classmates were applying is competing on leverage. Same degree. Different trajectory. The gap widens with every cohort. To me, AI doesn't just eliminate mediocre careers. It stretches the distribution. The top end pulls away faster than the middle can catch up. Totally agree to this, "the best time to enter is now, if you rise to the higher bar."

Ankur Pandey thought you would want to look at this

Ash Mohatt Exactly. When you are in the suit you won't fire randomly, NOT WITH SAFETY ON YOUR MIND!!!

Juliet MacDowell

Workflow integration is the real game-changer Paul Storm reducing tool friction beats chasing the smartest model.

One age is over. New is here 👏

Honestly if you know all the principles of how to code and you've spent 3 years studying it, you have the cheat code. You have the expertise to orchestrate AI systems and debug better than anyone else. And every business in every domain needs this right now and to become AI first. Lots of money to be made from this right now.

When the world evolves, we must adapt. This is not to say it will be easy or without consequences. Billionaires fund and develop ways to get richer and dominate. Unfortunately, people follow, adjust, lead or get out of the way.

@jerome enderlon

I think AI made us faster, but Stack Overflow made us sharper. AI is great for getting unstuck quickly, but the real skill is still knowing how to debug, verify, and understand the answer. The best developers will probably use AI like a senior assistant, not like a replacement for thinking.

And AI can "read" documentations that developers are lazy to do

The future may belong to integrated workflows, but the quality of the underlying model still matters. Less friction is great, but better outcomes matter even more.

Interesting point. The shift isn’t really about model quality anymore it’s about fragmentation of workflow

Love the analogy! Luís Rodrigues 👏🙌

Move this image upside down,,and that will be the position Muslims do more than 50 times, five times a day, that is called Sallah "Prayer."

your videos will never cease to amaze me. Love it Tianyu

Platform consolidation usually wins enterprise adoption because operational friction compounds faster than model capability improvements internally.

Michael James Folliett thanks!