frognumber

A quality mask costs around $1. Per-capita GDP is about $33 per day worldwide. Or put another way, with 8 billion people in the world, a mask a day runs around $3 trillion annually.

The world is interconnected.

I like masks since, aside from viruses, they filter pollution and allergens. I feel a lot better post-2020. However, universal HEPA filters and covid masks are very much for the rich.

1) If the exponent is >1, it doesn't make much difference. Everyone will catch "it," for whatever disease.

2) It makes sense to have interventions prioritized by ROI

3) Many interventions are individual, and there is no exponent. Should I, as a [doctor / immuno-compromised-individual / patient / etc.] wear a mask?

1) There are diminishing returns on investment, as we go further out. We'd need to go exponentially further out to get useful information.

2) The sensors on the craft are old, and modern methods can tell us more from far away then ancient sensors from deep space.

... and so on. I'm not going to argue about which assumption is best, because intuitions will differ. I was hoping to get a concrete answer from someone who might know rather than was speculating.

Ergo, the question.

I was hoping for an answer of the form "Nope. It's not worth throwing it away, but it's pretty obsolete now." or "Yup. We recently learned that ____" (fill in the blank).

Also, the people who originally developed X were brilliant and X was brilliant.

https://en.wikipedia.org/wiki/Survivorship_bias

In general:

- Brilliant people in the eighties will still write better code than typical people in the 2020s.

- Ditto for brilliant code. A lot of this is actually pretty random. Sometimes an idiot will architect something great by accident, and vice-versa.

- Most things written in the eighties still used are really, really good in some way (or they would have been replaced many times over).

So most attempts to replace classic technologies end up with dramatically inferior replacements. Classic Unix was really, really good. It survived not because OSes in the sixties and seventies were any good (most weren't), but because if you build many of them, one will turn out great.

I'm not sure there's much more than that.

I'll also mention: As the above comment implies, schools change over time. Reputations follow decades later.

One cannot legally practice law without a license. The definition of that varies by jurisdiction. Fortunately, in my jurisdiction, "practicing law" generally implies taking money, and it's very hard to get in trouble for practicing law without a license. However, my jurisdiction is a bit of an outlier here. Yours might differ.

In general, the line is drawn at the difference between providing legal information and legal advice.

Generic legal discussions, like this one, are generally not considered practicing law. Legal information is also okay. If I say "the definition of manslaughter is ...," or "USC ___ says ___," I'm generally in the clear.

Where the line is crossed is in interpreting law for a specific context. If I say "You committed manslaughter and not murder because of ____, which implies ____," or "You'd be breaking contract ____ because clause 5 says ____, and what you're doing is ____," that's legal advice.

The reasons cited for this are multifold, but include non-obvious ones, such as that clients will generally present their case from their perspective. A non-lawyer will be unlikely to have experience with what questions to ask to get a more objective view (or even if the client is objective, what information they might need to make a determination). Even if you are an expert in the law, it's very easy to accidentally give incorrect advice, which can have severe consequences.

In practice, most of this is protectionism. Bar associations act like a guild. Lawyers are mostly incompetent crooks, and most are not very qualified to provide legal advice either, but c'est la vie. If you've worked with corporate lawyers, this statement might come off as misguided, but the vast majority of lawyers are two-bit operations handling hit-and-runs, divorces, and similar.

In either case, it's helpful to give the disclaimer so you know I'm not a lawyer, and don't rely on anything I say. It's fine for casual conversation, but if tomorrow you want to start a startup which helps people with legal problems, talk to a qualified lawyer, and don't rely on a random internet post like this one.

On the low-end, I actually don't get how sellers stay in business with Amazon fees, but for the most part, my go-to is increasingly direct (Aliexpress, eBay, or similar).

I'm down to about two Amazon orders per month. That's down from several orders a week a half-decade ago. It's also a bit of an increasing death spiral, since I gave up Prime and Amazon is now just annoying without Prime.

I do feel there is room for a vendor to take Amazon's place. The next one should have:

- 100% direct supply chains. No knockoffs. If I order Sandisk, it either comes form Sandisk, or from Sandisk direct to the vendor.

- 2019-era Amazon customer service

- Fair prices. Fair doesn't mean beating Amazon and Walmart (customer service costs something), but that I never get hit with a 10x markup. I should be able to trust I'm not being cheated without shopping around.

- Honest reviews

- Ideally, over time, shoppers, vendor-recommended products, and expert reviews. Trader Joes and Apple to well by virtue of fewer, better SKUs. Buying dpreview could have been a great fit for Amazon, but as is, building out something like the old Wirecutter, Consumer Reports, etc.

- I can even see more social models, like forums with company reps, experts, etc. where I can ask a question before I buy.

The logistics here are a multi-billion-dollar play, but might be possible either leveraging local stores or in partnership with a Fedex or UPS.

However, it's worth noting the universe is a cesspool of corruption. If you pretend it works the way it ought to and not the way it does, you won't have a very good time or be very successful. The entire legal system is f-ed, and if you pretend it's anything else, you'll end up in prison or worse.

Programmers think like machines. Lawyers don't. A lot of confusion comes from this. To be clear, there are places where law is machine-like, but I believe licensing is not one of them.

If two licenses are substantively equivalent, a court is likely to rule that it's a-okay. One would most likely need to show a substantive difference to have a case.

IANAL, but this is based on one conversation with a law professor specializing in this stuff, so it's also not completely uninformed. But it matches up with what you wrote. If your history is right, the 2019 changes is where there would be a case.

The joyful part here is that there are 200 countries in the world, and in many, the 3.0->4.0 would be a valid complaint. I suspect this would not fly in most common law jurisdictions (British Empire), but it would be fine in many statutory law ones (e.g. France). In the internet age, you can be sued anywhere!

One of the upsides of doing a startup in parts of Eastern Europe, for example, is the ability to self-fund even for normal people. Cost-of-living is ultralow, since medicine is free, housing is cheap and available (at least outside of major cities), public transit is cheap and works, education works without private schools / tutoring, etc. Basically, $10k / year gets you the good life. And if you screw up, the safety net means there's not too far to fall; you land on your feet and try again.

That's a very favorable climate to bootstrap a business in.

You do see a lot of entrepreneurship from there of that type. It's not US-style big tech, but it works well. Your odds of building Google are low, but your odds of building a self-sustaining small-to-midsize business which you own are quite reasonable. The latter of those is likely a better thing to aim for.

If your goal is to raise $100M, give up 98% of equity to investors, and have the business explode or implode, Silicon Valley is the place to do it. Not everyone wants to get on that ride, though.

That's more a salesperson mantra than SWE, where for every 10-100 people you ask, someone buys something. However, I've raised money many times simply by shamelessly asking.

Second good life lesson: Don't assume things about others.

People who make $1/day, $10/day, $100/day, $1000/day, $10,000/day, etc. have fundamentally different priorities and motivations. Ditto on many other axes. Good synergies are leveraged working across such differences.

- Most of my career was made by being the author of one popular open source platform which happened to do well.

- I've recruited people based on open-source contributions. If I want an expert in [X], finding someone who contributed to [X] is a good bet.

- The flip side is I've made (minor, helpful) contributions to many projects in part for exposure. My name is in the commit list of many systems in domains where I have wanted to work.

- Many mid-sized contributions look good on a resume, especially for a junior developer. Indeed, I've made one case to promote someone based, in part, on contributing to a library we were using (even if only tangentially).

If you want a job in e.g. network security, find something in a firewall, anonymzing proxy, packet sniffer, or whatnot, and make a PR. It's often quick, easy, and helpful. A corollary is you do actually learn a lot about a system by contributing.

I have no axe to grind here, but I think the cynicism is unwarranted.

This would be an ideal place for a semi-technical user to contribute documentation. Good starter open-source contribution. Without that, it's hard to figure out if it would work for me, let alone give feedback.

They're no longer the cheapest option, and the salads are no longer available. There's something going wrong there. I can get cheaper, healthier, better food at e.g. my local Subway, and many other places.

It is still reasonably fast (although not super-fast). Unique to my location, the service is excellent.

In a few cases, this is simply fabricated, but in most cases, the techniques used don't raise to the level of data fabrication. There is a lot one can do in data analysis, interpretation, experimental design, etc. designed to lead to a specific result. MIT, at least, will not treat this as malpractice, even though the ultimate results are nonsense, and often lead to actual human harm when they drive real-world actions.

It pisses me off.

1) Early brains can rewire themselves. It's amazing. If you can detect dyslexia, autism, and similar early, you'll see changes on an MRI as it fixes itself with the right intervention.

2) This process is almost completely opaque. Teaching early math or reading guarantees kids will learn it later, but you see _almost no progress_ at first, and then zoom. You need to be patient. Most of the pieces the child is learning aren't the ones you expect (and we don't quite know what they are).

3) Brains are wired differently, and about 10% of kids cannot and will never learn to read fluently without explicit instruction in phonics. Sight reading / whole word does work for some kids, but not all kids. The "science of reading" is all about being explicit about phonics.[a]

[a] I use "science of reading" in quotes, since it's the name of a specific movement, and should not be confused with the common meaning. "Science" comes up in a lot of contexts from "Christian Science" to "Computer Science" with varied levels of science. "Science of reading," as a movement, is grounded in solid science, but a fairly narrow set of studies.

Cadance OnCloud Platform for PCB Design is $1500/month. Altium Designer is $11k.

Most of the universe is in the world of KiCAD, Eagle, GEDA, and similar.

The reason I gave a specific use-case is because in this use-case, it does work like that. I did not give a use-case which had RF signals, large currents, or much of anything else. I2C "fast mode" is 400 kbit/s.

The reason things like STEMMA work is because everything is slow enough that parasitics don't matter, and power is low enough that voltage doesn't sag. The lumped element model is nearly perfect there.

Even many domains not like that -- let's say audio -- have a lot of places where abstraction break, but where those failures are well-contained. An audio amplifier might have 50MHz gain-bandwidth, and the layout around it needs to be pretty tight, but once that's a block, things matter a lot less. That's why you can have all sorts of analogue synthesizers where people plug blocks together. A qualified engineer needs to design those blocks, but connecting them, abstractions hold pretty well.

No one will design an oscilloscope or a software-defined radio front-end using React for Circuits. However, as much as those occupy a lot of engineer time, they're actually in the very tiny minority of boards people want built. Most things are pretty simple and dumb.

Heck, even many high-performance designs have simple stuff around them. If you're making a radio receiver, the RF might be voodoo, but you know what's not voodoo?

- The controller for the buttons and the seven-segment displays

- The battery charger / monitor circuit (at least as a block)

- The USB-to-RF232 converter for development (at least as a block)

- The power, temperature, and other monitors

- The neato audio spectrum bar display

- The bass/treble filter (at least as a block)

- A lot of the test fixtures needed to test the RF equipment (not all; a lot of the other test fixtures are deep voodoo; it depends on what you're testing)

- Etc.

The core thing code allows is code, constraints, and subcomponents. I can see writing:

<oscillator555 frequency="10kHz">

And having this have a premade parametrized component with layout, as well as knowing about things like power supplies.

I can also see this integrating more with GUI tools.

There are many steps going forward, not specific to this, which would need to be in place, such as:

- Allowing React XML without JSX (e.g. in data files)

- Exporting to the above

- CSS-style templates and state management

... and similar. Those would potentially permit both code-based tools and GUI tools to interact meaningfully, and that's where the real power would come in.

You get there with prototypes like this one, and increments.

Ideally, where this would end up is almost:

<EPSHome> <ESP8266/> <Sensor/> <Sensor/> <Sensor/> </ESPHome>

And a crude-but-working layout comes out, which gets made for $10, populated with ESPHome, and monitors whatever I want to monitor at home. Not much more than STEMMA, but with a nice PCB instead of a dozen cables .

Aliexpress -> From $0.99 -> four-pack of ESP8266 for $3.19.

Even less when they have a sale.

If you're a shipyard, an oil company, a bank, an automaker, etc. you still need software to manage things like inventory, employees, logistics, and similar, and you have zero expertise to do it in-house. They also have zero expertise to find a qualified vendor.

IBM is a safe bet.

That's a huge market.

- Unsafe product is shipped

- Damage is done

- Someone pays for it (ultimately the consumer)

The difference is that if the consumer directly pays for it, companies have every incentive to make unsafe products, and those costs are very, very high.

If the companies pay for it, the free hand of capitalism steps in, and optimizes such that any safety measures where expected returns are positive are implemented (and those where expected returns are negative, not).

If a death costs $1M, and I can do a safety measure which reduces odds of death by 0.01% for $100, it's cost-neutral. If that same measure costs $50, an efficient business will implement it. If it costs $200, it won't.

That allows global optimization. By setting the price of death, you can have a set point for what safety measures will be taken, the system optimizes adequately well, and deaths go down, so fewer costs are passed back.

I can certainly think of a time when I was driving when I missed another car in my blind spot and almost caused an accident, or when I saw someone in a car distracted by a toddler, or similar. Those didn't lead to accidents, but they might have if someone were less lucky.

There's a range of ways a car can kill people, ranging from driving through a red light at 100MPH, to an illegal U-turn, to making a stupid mistake, to a completely random fluke of circumstance.

On one end of the spectrum, there should be prison time, and regardless of whether an accident happens. On the other end, insurance should pay damages, but I'm not sure what good the criminal system can do in terms of deterrence.

If a driver is already doing their best to be safe, but slips up, or even isn't doing their best but isn't being unreasonable, criminal penalties don't seem like the right outcome.

As much as it's absolutely the driver's fault, it's also an inevitable accident.

Wrongful death is about $500k-$1M (but should be higher). Tesla market cap is $500B. Damages for O(one million people) are equal to Tesla's market cap, so this is not even bad for business; it's simply factoring in externalities into the cost of doing business.

A good split might be criminal system for the driver (which ranges from fines to prison time, and is designed to be punitive), and civil system for Tesla (which is only financial, but designed to cover the financial cost of the death to the family).

If history is any indicator, he seems strongest at mid-stage companies. He didn't start Tesla. He isn't doing well there right now (or Twitter, for that matter). He's too erratic for a mature company.

But he did an awesome job taking Tesla from the $M to the $B stage.

Any company at that stage would also benefit from the visibility he brings. For smaller players, any press is good press.

I think the other question is how likely he would be to walk away. I think at $0, Musk would likely stick it out at Tesla, since so much of his net worth, ego, identity, and brand is locked up in there. In other words, this seems like a bluff.

Google would need to be, say, half the cost of AWS to get my business.

Toss in hiring head of product from Oracle? No one in their right mind would use Google Cloud