HN user
Headlines like this are more pleasant than headlines about how the $100M project to implement SAP is five years behind schedule, how Deloitte has consumed tens of millions of dollars that could have gone to patient care, or that the new ERP system has lead to a total lack of accountability for patient outcomes.
Needs a mode where touching snow counts.
Probably by trying to do it from first principles.
Humans are actually intuitively log scale thinkers. That is, humans without the kind of early arithmetic training that Westerners get will think more in terms of ratios than differences. There are theories it is more evolutionarily adaptive.
Sounds like a Greg Egan writing prompt.
I’d imagine longer term contracts not renewing or renewing at lower prices / lower volumes.
But can their AI quietly introduce a security exploit into a GitHub project?
I dunno, the new Microsoft / OpenAI computer is named “Stargate,” that’s a pretty solid name.
I was disappointed that the lab assistant is not named.
I’m disappointed it has nothing to do with the logo programming language.
Reusing short English words is a big problem in computer science.
Love seeing real time streaming connecting through to visualization. You make Grafana look really good. Did you run into any limitations there?
Many years ago I had a job interview that included a personality test. I didn’t take the job, for other reasons. But I always wondered what had gone wrong in the past such that a personality test for engineers was in order. I’m now glad to know. I also wonder how many problematic malcontents I would have found had I gone to work there.
Premature denomalization is expensive complexity. Denormalization is a great tool, maybe an under-used tool. But you should wait until there are hot paths before using it.
I think of it as a bunch of bad early implementations that are being used for dubious financial purposes and outright fraud, fueling a bubble. The same could be said for joint stock corporations in the 1600s. Something can be inevitable (maybe it is maybe it isn’t) and the current situation can still be bad.
Someone in the 1600s saying that joint stock companies are terrible, solve a problem that doesn’t really need solving, and are being used to rip off investors and fund unethical colonialism would have been right. It took a while before people used them in good ways. And even today they are questionable.
I don’t think we need to solve non-problems in the interest of foolish consistency.
I canceled recently and had to escalate to “I have told you to cancel three times and if you cannot cancel now I will have to dispute the charges with my credit card company.”
Can confirm, my kids get letters at camp from their grandmother written in cursive and they basically cannot read them. They can puzzle out cursive when it is very neatly written, but casual cursive handwriting is frequently impenetrable.
I remember the days of going down to my friends store because he had a modem and knew which BBSes would have software downloads. I had to bring my own floppy disk because those things cost money.
I dunno, I watched two professional architects not be able to estimate a beam size for a 16’ span without mocking it up in autocad. This delayed us like a week in design iterations, and cost me probably $2000 in architect hours. I own a book inherited from my grandfather that just told me the answer. A 72-year-old architect who helped with a later revision of the project also knew the answer off the top of his head.
That’s not to say I would build the house without doing a proper structural engineering analysis and simulation. But not being able to estimate feasibility without doing the full simulation is a real problem.
I like to flip a coin when the decision is difficult. If the answer is satisfying, then either I had an unarticulated preference or it didn’t matter. If the answer is unsatisfying, then I had an unarticulated preference and I can just choose the (apparently preferable) alternative.
The causal connection can be remarkably tenuous, especially if you add securitization or other financialization into the mix. The people supplying the money often care more about hitting standards than maximizing overall eventual profit. Understanding this aspect of human and market behavior is useful for realizing why a lot of libertarian ideas are unworkable in practice.
There are already a handful of Lorawan companies that have built out networks or partnered with telcos. They have raised money from places like Comcast. They will probably exit back to places like Comcast. Governments are often their customers. And they are way too boring for a16z.
I did one once and it was a huge PITA and a very slow timeline. The free money part was nice, but the overhead and timeline mean I’ll probably stick to other funding sources in the future. If your founding team has extra PhDs and you want to pay them to do “research” it might work.
Finding a big problem is actually optional, and often a distraction. Most companies that are already functioning at scale have plenty of medium sized problems that need strategic fixes. For example “sales aren’t growing in Japan.” Most small companies are happy to stay small and also need only solve medium sized problems. For example “there is no good construction supplies vendor in this geography.” It’s only startups looking for rapid-scaling venture-friendly returns who need a big unsolved problem they can grow into.
Of course, those companies get a whole lot of attention. But Google and Facebook and Twitter don’t need big new unsolved problems. They already picked their big problem, and solved it well enough to get big. Now they need strategies which address the medium sized problems that limit their continued success. When they go looking for big new unsolved problems (I’m looking at you Alphabet and Meta) it’s usually a distraction.
The “shadow foster system” is related to this and has some problems. These are complex situations, maybe sometimes there is no good options. But if you are interested in learning more, Propublica has done some good reporting on the downside of non-foster placements: https://www.propublica.org/article/they-took-us-away-from-ea...
I’m all for questioning a group’s motives. But protecting personal liberty for everyone sometimes requires limiting personal liberty. I like to be free to not be killed by a drunk driver, for example, and that means we cannot leave drunk driving up to “personal responsibility.” I generally think people should be free to not be assaulted on the street, or by their spouse, or by their parents. I’m not sure that restricting alcohol is the best way to do that. But if it were I don’t think a principle of individual responsibility would get in the way. The crimes they are discussing are not victimless, they impact both individuals and society.
https://en.wikipedia.org/wiki/Mouse_Systems became available in 1982. My first mouse (on an Apple 2c) came with drawing software included.
I’m not an expert, but 20 years ago I had a small part in DARPA research into monitoring soldiers’ hydration. The idea was that people in combat often don’t notice they are thirsty, or power through it, and as a result are operating at reduced cognitive and physical capacity. Of course, that’s combat. And for all I know the funding was based on the popular misunderstanding.
This is sort of liking saying “how hard could it really be to do a special software build for one customer?”