knuthsat

27 ms·

241.
by knuthsat 6y ago
Tools for exploring the models are already being developed. Language models that are not given concepts of part-of-speech, dependency trees, grammars, develop representations of their own and people have found ways to inspect them.
242.
by knuthsat 6y ago
But let's "profile" how bad they are. Comparing Bitcoin to Argentina is not good profiling. Also, optimizing Bitcoin probably makes less sense than optimizing some of the other offenders in energy use and pollution. Why are w
243.
by knuthsat 6y ago
Well, I'm pretty sure Bitcoin does not have livestock waste, air pollution and similar impact compared to the energy use of Argentina. So comparing Argentina and Bitcoin with just energy use is a bit odd to me too.
244.
by knuthsat 6y ago
This article is just bad. There is nothing wrong with proof of work. It is useful. People also use Bitcoin, Ethereum and other cryptocurrencies for useful things. Energy demands are nothing compared to other useful human activities. Also, e
245.
by knuthsat 6y ago
You might find zero suppressed binary decision diagrams (ZDDs) useful. They also allow simple random generation of all kinds of objects (even constrained ones). Knuth adjusted his algorithm X for ZDD and got some nice speedups.
246.
by knuthsat 6y ago
I'm aware of approaches similar to this: https://gist.github.com/DavidBurkett/32e33835b03f9101666690b... describing offline transactions.
247.
by knuthsat 6y ago
Unfortunately, there really aren't that many better. Full transaction privacy (addresses not visible to public + amounts not visible to public) does not exist in any of the top 10. + offline transactions should be possible (allowing sy
248.
by knuthsat 6y ago
Maybe better to check unscented Kalman filter instead, because the extended one is a mathematical horror.
249.
by knuthsat 6y ago
KF algorithm when layed out as: 1. Figure out the initial state 2. use process model to predict state at next time step 3. adjust belief of the true state by taking account prediction uncertainty 4. get a measurement and belief according to
250.
by knuthsat 6y ago
It would be nice if the market is the only force creating demand for production. The amount of subsidies is massive for soy/corn <-> cows/poultry production. Removing the subsidies would make burgers and meat in general muc
251.
by knuthsat 6y ago
let's say you have a newline delimited dates in &junk. now you do a while loop, whoops quadratic.
252.
by knuthsat 6y ago
Depends on how much time (in hours) you spend working. 2 years is maybe 1 year of working time.
253.
by knuthsat 6y ago
not much.
254.
by knuthsat 6y ago
That JSX element is a function call, not a class that has a render method. If that <ExpensiveComponent /> was a `class ExpensiveComponent extends React.PureComponent { render() { ... } }` it would not be an issue because diff alg
255.
by knuthsat 6y ago
If expensive component extended React.PureComponent there would be no rerender.
256.
by knuthsat 6y ago
If expensive component extended React.PureComponent instead of being just a function call it would still work. New appearance of hooks and effects made people not use PureComponent class and replaced everything with functions that return JS
257.
by knuthsat 6y ago
Yeah, what I dislike about Upwork is that after they take thousands of dollars from me for their service, they still want me to pay for applying for new work. I’m the one that spends time filtering through their cluttered job list, I’m the
258.
by knuthsat 6y ago
There's no big sugar behind it. You can quite easily see that a bunch of studies are on isolates, the Japanese he mentions eat much less calories and more fructose as a percentage of calories, etc. There's just too much inaccuraci
259.
by knuthsat 6y ago
Eating the equivalent amount of apples to HFCS is never going to have the same effect. Fiber content and other good stuff will completely remove the effects that you might see fructose isolate.
260.
by knuthsat 6y ago
Those videos have been debunked. Lustig used a lot of false claims.
261.
by knuthsat 6y ago
I think there's not enough researchers that publish code. For example, discrete optimization research (nurse rostering, travelling salesman, vehicle routing problem, etc.) is filled with papers where people are evaluating their methods
262.
by knuthsat 6y ago
I had the same horror dream from about 2 years old to 12. It repeated before sleep walking or when I had fever. It brought the same feelings and had the same motifs every time, and it was looping (I'd get massively scared and the dream
263.
by knuthsat 6y ago
Energy is linked to price. Not to transactions.
264.
by knuthsat 6y ago
We had a similar assignment in elementary school and for some reason the memory of teacher being surprised with the small number of species we found on the grass around school is still in my mind 15+ years. I was so convinced there was noth
265.
by knuthsat 6y ago
You're assuming the rest of the world is different but it isn't. There's a massive amount of companies where working pace is very slow. A large percentage of employees do barely anything and these companies exist forever.
266.
by knuthsat 6y ago
Early return ifs are great. I usually just transform functions with 2+ if statements into functions without `if` statements by extracting `if` statements into other smaller functions where I can write it with an early return. Looks great.
267.
by knuthsat 6y ago
When I look at Knuth code snippets in CWEB I get the same vibes. He has a full implementation of garbage collecting zero suppressed binary decision diagrams that he used to count various stuff and it’s insanely elegant and minimal.
268.
by knuthsat 6y ago
I’m currently in exactly that situation. No comments on tiny workarounds, no high level docs on any feature, any kind of code criticism interpreted as personal attacks. I’ve worked at teams where I can easily add 1000+ lines of well tested
269.
by knuthsat 6y ago
Knuth reduced it to half a page in the comments
270.
by knuthsat 6y ago
The world is irrational in doing 0.x% improvements on something with a small footprint. I'm waiting for a time where programmers will find it completely rational to optimize the fastest functions in their code. This whole environmental

More ›