kva

The AI industry is monofocused on difficulty over utility. The NS problem is hard, but it is also useful if solved in a way that it illuminates a lot of areas of math, gives us new techniques etc.

Brute forcing a counter solve has not really changed the game as much.

Rhythm | Senior Embedded/Firmware Engineer | Chicago, IL | ONSITE (Hybrid ok) | $180k–$250k + equity

Rhythm is a medical device startup headed toward FDA clearance. We're hiring a senior firmware engineer to own solution architecture and embedded development: C/C++ on ARM Cortex-M, BLE, sensor integration, IEC 62304 through 510(k).

Looking for 7+ years of embedded experience, time in medical devices, and startup experience

Email: kanishk@rhythmpap.com

Big big fan of LTSE; The core tension I point out is simply that the epistemic crux of TLS -- short term validation -- is a poor proxy for the core risks in frontier companies. The cultural result of that is most people refuse to start long term companies since there is no short term way to validate the demand, which may in turn reduce demand for the LTSE itself :)

Looking forward to reading Incorruptible.

The lean startup methodology is fundamentally antithetical to companies like OpenAI, Anthropic, Antares, etc.

How do you resolve the difference between the short term nature of the lean startup, and the long term optimistic nature of LTSE?

Rhythm | Chicago / Remote (Ok for exceptional candidates) | Senior ML Engineer | Full-time

We are a KP-backed startup that is solving sleep apnea for 1bn people. We are looking for an exceptional ML/AI engineer to help build human simulators.

Our team is full of world class technical talent and this is a once in a lifetime opportunity to meaningfully improve the daily lives of people around the world.

Please email us a proof of any one exceptional AI project at kanishk [at] rhythmpap [dot] com.

Bonus points if you've had to titrate a CPAP before

Totally agree - most of these co's that get funded are pointless. FWIW, the general math here is that you'd spend < 2-3mm developing new git and most of the money goes into distribution.

We've strayed really far from where technical innovation began

RhythmPAP | Full stack hardware engineer | Full-Time | Chicago

This is your chance to work on a product that will meaningfully improve 1 billion lives. Sleep Apnea is a terrible disease that has no good treatment -- our new treatment has proven to be very effective in preclinical trials. Users rave about our product, and it has received a lot of love on social media. We are backed by tier 1 investors and angels.

We're hiring a founding hardware engineer that will help refine our prototype with an industrial designer on staff.

We're offering 200K+ / yr and 1%+ equity for this role.

to apply: email kanishk @ rhythmpap . com

[1] https://x.com/reggylong/status/1966209250968944942

UARS = upper airway resistance syndrome RERA = respiratory effort related arousal

UARS is effectively a "cousin" of sleep apnea which tends to occur in younger and fitter people - it is hard to diagnose but quite debilitating. RERAs are the primary sleep disturbing event for folks with UARS. Similar to how apneas are the primary event for folks with sleep apnea.

Hello! I am one of three working on a new treatment for sensitive sleep apnea patients / UARS Patients. We'd love to hear about your experience with trad CPAP and OSA.

This Or "Learning Curve Pricing" which dominated the outcomes of early SV.

someone needs to bring up the indestructable Nokia

Check out the WatchPAT One - its a new at-home sleep test that is really simple and very non-invasive.

This just represents such a stark lack in originality and lends even more credo to "OpenAI wants to copy its customers"

Hopefully Martin Casado or one of the other awesome open source folks from a16z will take a look at this and make the person whole!

A somewhat niche but good book on sleep-disordered breathing is "Life Saving Sleep" by Barry Krakow [1].

If you have a ton of sleep problems, you should probably catalog them, and try seeing a somnologist or getting an in-lab sleep study.

[1](https://www.amazon.com/Life-Saving-Sleep-Horizons-Treatment/...)

is there anything that VCs will _not_ pontificate on?

Unfortunately, there is no textbook on something very particular like "The CPAP Market in the US" - which is what I want to understand currently.

Also textbooks frequently miss out on the level of understanding part of this. There is no textbook that explains quantum mechanics for smart 15 year olds.

Given the right prompt, I'm sure it is....but when do users ever enter the right prompt? :(

Dehydration really is the one that gets me

What tools does your son use?

You might be slightly depressed but that's not everything. The truth is that most software companies are...pointless. They're sales driven body shops talking about "mission".

I really like my job - work on innovative product lines with a small close knit team. I got here by just working on ideas/probelms that interested me and selling that company to the current one.

Go do something you like. Also go to a therapist.

I <3 GQL but this just sounds like "Forget the GraphQL DSL to learn our DSL"