haskellandchill

I agree. Very shallow "sources" and no qualifications to be speaking on a medical topic. There is a role for pop science to disseminate knowledge but this is pure speculation.

The current situation is that you have to go out of your way with things like `pip install --only-binary`. There is a lot of implicit trust in developer tooling.

I think it is common that in installing packages you have hooks to execute code anyway.

I thought it was a good write up but did anyone catch the ending comment?

> We believe artificial general intelligence will be created within our lifetimes, and likely within the next decade.

Maybe within our lifetimes (if you are young) but I find it highly unlikely within the next decade.

tech to med school pipeline, don't recommend it but I'm happy

proud owner of a financed 2024 manual nissan versa here :) but yeah the dealership made almost no money and I put down a deposit when it was a couple months from coming in at a location far from where I live. it's a $20k car though.

It's not just one unpaid intern with long term memory loss, it's several of them. And they don't need breaks.

If you could hire an army of unpaid interns with long term memory loss who work 24/7, would you?

From your link

> Many studies have shown the incidence of repair procedures and worse final vision outcomes were higher in groups with autoimmune conditions (SJS, OCP). The difference in outcomes appears to be related to the degree and cumulative past period of inflammation. Overall most favorable outcomes are achieved in non-cicatrizing conditions, followed by ocular burns and OCP with the worst outcomes in SJS patients.

The patient in the article was a SJS patient

> The massage therapist says he could see just fine until he was 13 years old, when he took some ibuprofen after a school basketball game, triggering a rare auto-immune reaction known as Stevens-Johnson syndrome.

For those without means understanding is a luxury, for many with, it is merely aesthetic. Matching drive with opportunity could unlock humanistic discovery but it is far more likely to be done artificially given the way we organize our societies.

I gave up a while ago and started on the path to becoming a doctor. I'm pretty happy about it except for the 10 years until I make what I used to part.

Not really. I had projects like that at Georgia Tech, I know Carnegie Mellon, MIT, Stanford, etc have similar.

Active learning and problem selection is what I came to as well going through the literature. I was thinking of teaching algorithm design by having students build their own algorithm laboratory where they create visualizations and experiments that motivate interest in problems and the design of solutions.

Hm this seems potentially dangerous, doing MCAT practice I got:

  Which of the following is NOT a neurotransmitter in the central nervous system?

  serotonin
  glutamate
  acetylcholine
  dopamine

  Neurotransmitters in the central nervous system include serotonin, glutamate, and dopamine. Acetylcholine is a neurotransmitter in the peripheral nervous system.

However this is wrong. Acetylcholine is both a PNS and CNS neurotransmitter. The rest are also CNS neurotransmitters.

Here is an example of a real question that is similar but accurate: https://quizlet.com/explanations/questions/which-of-the-foll...

You can say whatever you want, there are studies backing up the beneficial impact of quantitative over qualitative.

It’s really not unquantifiable. I read “How to Measure Anything in Cybersecurity Risk” and it was an eye opener. Using a table of risks and outcomes with associated probabilities and 90% confidence intervals of dollar impacts we can quantify categories of technical debt.

The problem is the muda should have expected values associated with them. Bugs and security vulnerabilities do cost money, these are 90% confidence intervals of dollar impact from How to Measure.

Here is a blog post: https://www.linkedin.com/pulse/three-project-management-meas...

What about the Bayesian methods shown in "How to Measure Anything"? They have been applied to Cybersecurity ("How to Measure Anything in Cybersecurity Risk" in a very thorough and convincing manner. It looks like the business around it is trying to apply it to product management (https://hubbardresearch.com/shop/measure-anything-project-ma...). Basically the idea is when things are hard to measure we should not abandon quantitative scales and use qualitative ones (like t-shirt sizes) but instead use probabilities to quantify our uncertainty and leverage techniques like bayesian updates, confidence intervals, and monte carlo simulations.

It's not much better if you get responses. My problem was companies will schedule me for interviews so I do 100 applications have 60 phone screens, then 40 technical interviews, then 30 seconds stages/homeworks, 10 final rounds and maybe 4 calls with the VP/CTO and still nothing. It's a full time job for months with no payoff.

that sounds like a great open source project, I'm interested in starting on it if there's nothing out there.

I'm kind of glad it's a dumpster fire right now. If I got a job like I had in the past I wouldn't have decided to go down a different path. I'm going back to school to complete premed requirements and apply to medical school next year. I'm working as a medical assistant and training for EMT. I'm much happier than I would be as a web developer. There are tech jobs that would have interested me but I'd still rather be a doctor anyway, not that that is guaranteed. I did some open source Rust and learned some AI but you really have to show devotion through unpaid work to secure those positions, or get lucky.

Now frame this in type theory, zippers? The derivative of a data type is its one holed contexts?

Or is it not amenable to that perspective all I can find is from a different viewpoint: “ The dynamic trees problem, first posed by Sleator and Tarjan is to maintain a forest of trees subject to the insertion and deletion of edges, also known as links and cuts. Dynamic trees are used as a building block in a multitude of applications, including maximum flows, dynamic connectivity and minimum spanning trees, and minimum cuts, making them a fruitful line of work with a rich history.”

I’m interested in how this compares to the coalgebra approach like in “Session Coalgebras: A Coalgebraic View on Session Types and Communication Protocols” (https://arxiv.org/pdf/2011.05712.pdf).

It does seem all the interviews I've been doing are a waste of time. Even when I make it past the final round the company waits a week or two and says well you did good but we have too many qualified people or something.

I've been trying pretty hard for 4 months now making it the longest I've ever been unable to get a job. I have until end of March so it's looking dicey. Had to file unemployment.