cactusfrog

We have robust cloning protocols for dogs. For some reason dogs are really amenable to cloning.

The joke is that his greed/ unwillingness to squeeze margins has made the entire database company ecosystem possible.

Every waste of money is not a Ponzi scheme.

It’s a closed source POS with unpredictable performance. Talk to people who have actually been forced to use it and many have terrible experiences.

https://news.ycombinator.com/item?id=16170955

I’m glad they are wasting Walmart’s time and money.

It isn’t a suitable language for data science and lacks the ecosystem of Python.

For example, people will write Eden files to parse data, which involves creating a really complex DSL for every file for import.

They will try to use Datomic as a database, which requires using datalog and does not scale.

The worst part is the clojure users are convinced it is somehow better, and ignore the business use case.

I feel bad for companies because it clearly isn’t a good fit and those devs will hire all their friends and do nothing for five years.

This happened at an organization doing cancer research, so the selfish cultists ruined a chance at understanding immunotherapy by focusing on writing clojure rather than doing science.

Recursion pharmaceuticals also had this problem, and had to rip out all of the clojure from their system.

Why isn’t the input a csv?

I just did this in a few lines of python with numpy. I don’t understand why the input isn’t a csv.

This is a gamblers ruin problem action.

I’m working in biotech. I think creating “curated” materialized views with dbt + datawarehouse or duck db + using webgl with d3 will be best long term solution https://blog.scottlogic.com/2020/05/01/rendering-one-million... .

Matplotlib works well for static plots. Altair and others freeze at around 4000 data points, which is crazy. Streamlit + matplotlib is impossible to maintain but is quick to get up and running.

Interactivity is not desirable in all situations. For static figures, matplotlib is still king.

Twitter/X keeps spamming me with NSFW things so I deleted it

A significant amount of health issues result from medical noncompliance. A digital twin might be useful for finding ways of manipulating patients into taking care of their health.

I think it has correctness issues

Just write a preprint or a blog post

Probably promote cancer cell growth as well

Bone density loss in space is estimated at a loss of 1-2% of total bone density per month of microgravity exposure. The worst cases of terrestrial bone loss are around 5% of total bone mineral density per year.

Yes, this is one of the major problems with this. I don’t like pdfs

Biochemical engineering exists as a discipline and focused on “scale up” of production problems like these. Robots are involved, but most of the time the process is modified to a more stable one.

“There is an immense diversity, but also disparities in skill, expertise, and knowledge among Data Scientists. The difference in skills between a data scientist with - say - a Bachelor in Physics and a data scientist with a PhD in Computer Science is comparable to the difference in competence between a certified nursing assistant and a surgeon”

My experience, is that the longer someone works in academia, the worse of a coder they are. I’d rather have someone with 5 years of industry experience over someone fresh out of a PhD program.

The author has a PhD and is trying to impose the flawed hierarchy of academia onto tech. This push should be resisted.

The Mayo Clinic sounds like a garbage healthcare provider.

You have to pay enough to convince everyone to sell. The trading price is the marginal price of buying one more unit of stock, not a representation of the stock price at which every owner will sell.

This is not true, but it is a major challenge. See https://www.pathai.com/

Yeah. One application that I’ve heard is retrofitting indoor animal farms to capture methane and reduce methane emissions.

Should carbon capture should be worked in secret because mentioning it delivers false hope?

Gas separation has many applications.

If Google developed an AI model for cancer detection in MRI scans, you wouldn’t claim that “AI is not for what people think it is for” because Google also uses AI models for advertising.

I don’t believe I’m exaggerating. You can replace “Clojure” with “Jesus” in the linked article and it still makes sense.

The Clojure community is extremely cult like. I worked with a dev who tried to push Datomic, and despite the huge flaws with his system, and the fact that it was closed source, he would continuously tout how much better it was than a relational database. I don’t believe the hype.

People seeking to maximize profit are not in favor of expensive clinical trials and are not lying about how expensive they are.

Apple started out as a personal computer company—which was a radical concept because there was no good answer to “What are regular people going to do with one”. Mainframes and terminals were useful, but personal computers only had minor applications.

The Vision Pro product is inline with this philosophy of building a good product with no clear purpose and letting users figure it out. It is going to take years for someone to come up with a good use. Think about how long it took AirPods and Apple play to catch on.