cactusfrog

The jobs robots replace aren’t good for the human soul

I’m here and grew up here but I don’t really have friends or belong to a startup ecosystem. I just do my job and then go home. Idk how to take advantage of the area.

I think “copy from a PDF” could be improved with AI. It’s been 30 years and I still get new lines in the middle of sentences when I try to copy from one.

I believe LLMs could be useful here as a component of pre-commit hook.

Interesting, thanks for the comment. I didn’t know this when I posted.

Will I be able to change the space I’m in based on my mood? It would be fun if I could manipulate the algorithm serving me content in real time. For example, I would like to turn on “make me sad” mode or “make me want to exercise” mode.

Wow, clojure devs are learning that ides are useful

Even the boundary for newtons approximation is fractal. This is a feature of non-linear optimization.

They are not likely to be mixing powders and instead perform serial dilutions of the suspensions.

Industrial food preparation steps can be patented but they probably cannot be copyrighted.

I’ve used Jupyter notebooks with deno decently. They seem much better than Python Jupyter notebooks because of the lack of pain around dependency management.

Software devs are moving to other industries, which desperately need them.

I hate change and wish I could have a change card.

It’s really useful for learning. “Describe this complex bash command I found on stack overflow line-by-line”. Write a quiz question to make sure I understand it. Rate my response to this quiz question.

What do you think about clojure?

What’s strange about this case is that there are lot of carcinogens and evil pesticides out there that are not heavily regulated. For example processed meat has a very strong link with cancer and the nitrites added to processed meat are likely the culprit. No one seems to care and no one is suing the meat companies. There are “export only” pesticides that are truly scary that contaminate the us food supply and cause birth defects and cancer for entire villages. Roundup isn’t mechanistically tied to cancer and there isn’t a scientific consensus surrounding if/how carcinogenic it is. I wouldn’t be surprised if your brother’s cancer was environmentally related but I would bet that a boring chemical like gasoline/exhaust would be the cause. Gasoline and exhaust products are known to be carcinogenic.

We are using a Datomic derivative at my work because Postgres doesn’t have this feature. I don’t think anyone has actually used Daromic’s capabilities for this purpose.

Reading smresearch papers on BART is really annoying right now so I’m a fan of this project.

I think a major benefit of open source is that it doesn’t require good intentions on the part of the maintainers. A consortium of maintainers ensures that none of them have the market power to change the license.

I don’t see any value in these principles.

Minimal Viable Products over prototypes

What’s the difference?

APIs over databases

I feel like this is a terrible decision that has nothing to do with data science.

Clever use of computation over convenient assumptions

Why not both?

Dashboards over reports

Dislike this as well. Dashboards don’t contain analysis.

Validation, scrutiny and repeatability over convention and ad verecundiam

Sure, but this is not controversial

This is all from memory: Diethyl-ammonia works but is energy intensive to regenerate. Zeolites require less energy to regenerate but also absorb water. Some zeolite crystal structures might exist that are less hydroscopic, but have yet to be discovered. MOFs are like zeolites, but easier to fine tune and some good CO2 absorbers that are also not too hydroscopic have been identified. MOFs have some serious stability issues.

I’ve been the only US citizen on a team and now I am on a team with exclusively white, remote employees. I’ve watched my current team build the worst systems I’ve ever seen while constantly congregating themselves.

Null means unknown. I think the main issue is that people use it to mean “not applicable to this row”, which is different than null’s initial purpose. I think the problems with null could be solved by adding in another None value to distinguish between these two situations.

In the beginning of the 1920s, the leading American oil and automobile industry add lead to gasoline to get a more efficient fuel. At that time, lead was already a known poison, harmful to the nervous system - especially to children. Quickly, lead outcompetes other harmless additives, and the profit of the oil industry is huge in the decades to come. But hundred thousands tons of lead are spread around the world and cause injuries on millions of people - among others in lost IQ. Lead from gasoline has, literally, made a lot of us a bit more dim.

I’m working as a data science as at a biotech company, but I have a chemical engineering degree.

I want to highlight how cool bioreactor process optimization is. There is so much complexity and coolness to the problem. For example, scaling up does not involve making all of the parameters bigger because of different scaling relationships.

The issue in CA is that everywhere is the suburbs. My grandma’s house is in a single family neighborhood 10 minutes away from a huge office building with Illumina on the outside. The fact that this company, which has revolutionized biological research is in an area without enough housing and requires its workers to drive 1-2 h through traffic is insane from a society development perspective.

Suburbs can still exist but you need to have density around transit and provide an alternative to sitting in traffic.

How would it be terrible for you?

California property taxes are fixed to when you bought the property + 2% change per year

The good news is that no BMW is going to run for 8 years

I feel like 10x developers are those that spent 2 years working on the same problem at their last job and now know a solution