adi4213

> There's just not a lot of juice left to squeeze for Gemini to tell you exactly how tall Ke$ha is or when the last time Brittney Spears went to jail was...

But there is a ton of juice left to squeeze when it comes to post-training/RL for a ton of useful things in practice, right? It’s been amazing seeing how good modern model tool use is for example, and I bet there is a lot of room for improvement still (no doubt that a ton of improvement can be made more easily on the agent harness front or via post-training regimes like LoRa (which does support to your point about diminishing pre-training juice))

Can you provide some info on how to use the max plan / ChatGPT pro subscription via OAuth? I was about to subscribe but wasn't entirely sure how to configure that

Woah - I just drove by the JCV institute in La Jolla yesterday and felt thinking that the organization was really monumental to synthetic biology. Rest in peace

Can you use a ChatGPT Pro or Claude Max subscription for this?

This is really interesting, congrats on the launch. The use case I’m trying to solve for is building a coding agent platform that reliably sets up our development stack well. Few questions! In my case, I’m trying to build a one-shot coding agent platform that nicely spins up a docker-in-docker Supabase environment, runs a NextJS app, and durably listens to CI and iterates.

1) Can I use this with my ChatGPT pro or Claude max subscription? 2)

This demo started out as really cool - but the very first generated was terrible (in that it actually would have been very negative for our brand) Happy to discuss what exactly was problematic with you guys - so please take my comment constructively.

You should try www.betterbrain.com/insurance! It covers this set of bloodwork and 92% of covered patients pay $0

This is really interesting! Pretty remarkable that the key insight isn't just about crossing the BBB—it's about repairing the transport machinery itself.

We're building BetterBrain around this insight—catching vascular and metabolic risk factors that compromise BBB integrity years before severe pathology. Not replacing breakthrough therapies like this, but complementing them by intervening earlier in the cascade. (www.betterbrain.com/insurance if anyone would like to try - its $0 with most insurance plans)

Congrats on the fundraise! I've been a fan of Counsel and admire Muthu's bullishness around asynchronous care delivery.

Overall, I felt the human-clinician side of the product worked quite well - the providers responded quite quickly, were well informed, and resulted in a quick turnaround to getting an Rx or some manner of outcome.

On the AI side, I would caution that your team really needs to work quite hard to maintain a defensible moat here.

> Other chatbots can’t connect to your medical records to provide personalized advice. They’re also not trained on the latest evidence-based research and can’t bring a physician into the conversation. With Counsel, you get medical-grade AI support backed by physician oversight, ensuring safe, high-quality care

"can’t connect to your medical records to provide personalized advice" - while I think Counsel's HIE integrations seem to work well (the AI responded with relevant data it had on me from the exchange), that's is reasonably easy to build (which does sound like the famous Dropbox comment).

"They’re also not trained on the latest evidence-based research" - on this note, I'd be curious if Counsel is capable of doing meaningful post-training to meet this endpoint (I'd assume it's not worth the resources to explore any manner of post-training). It strikes me as a good example of Sam Altman's steamrolling comment - I'd imagine the next waive of frontier models will represent a much greater level of health inference than post-training a current frontier model will accomplish.

tl;dr - I found myself immediately skipping over the AI functionality but did find the human physicians on Counsel's app to work great.

Granted, I'm curious to see the company's direction long-term — perhaps along the lines of a16z's healthcare superstaffing thesis (which sounds compelling) or aiming to have most encounters be between a user and AI (which I'm more bearish about, though it's definitely going to happen one day, of course).

I think the key is remind yourself is that an engineer is supposed to solve business problems. So use these new tools to be more effective in doing so. An analogy is that people used to spend tons of time building out web server code but something like Django added tremendously useful abstractions and patterns to doing so, which allowed people to more productively add business value

This is very much part of the motivation behind BetterBrain, the wait for an in-network neurologist is often 6+ months, but the wait for one of the few in-person dementia prevention clinics can be years (and cost thousands out of pocket). If your insurance plan covers BetterBrain, definitely recommend giving our insurance covered option a try to meet with a clinician as soon as next week. We do have a free option, with the ability to purchase bloodwork and additional testing.

Yes, there are indeed paid options! We regularly support clients who travel to NY or Vermont for bloodwork. Our co-founder is Canadian and we work with many clients from Canada. Could you shoot me an email and our team can help figure out the logistics? [email protected] (Happy to send you a discount code)