In sparse coding, you're generally using an over-complete set of vectors which decompose the data into sparse activations.
So, if you have a dataset of hundred dimensional vectors, you want to find a set of vectors where each vector is well described as a combination of ~4 of the "basis" vectors.
Flaky tests: tests should be deterministic. If your tests are flakey in a 100% controlled environment, probably your real system is unreliable too.
Changing business requirements: business logic should be tested separately. It is expected to change, so if all of your tests include it, then yes of course it will be hard to maintain.
Too many tests for the same thing: yeah then maybe delete some of the duplicates?
Taking too long: mock stuff out. Also, maybe reconsider some architectural decisions you made, if your tests take too long it's probably going to bother your customers with slow behaviour too.
I've heard the same excuses from ML engineers before introducing tests there, embedded engineers, robotics engineers, systems engineers, everyone has a reason.
The real reason? It's because writing tests is a different skill and they don't actually know how to do it.
Oh that's crap. I've been a software engineer for over 30 years. I love tests - I preach testing at my current place of work. I've also worked in games for about a decade. Testing in games is... not useless, but very much less useful than it is in general software engineering.
I have no experience in the gamedev industry. But based on the number of bugs I see in games, plus the size and quantity of post-release patches, maybe your perspective here is because you're not trying to hit a level of reliability at launch that would justify having more tests?
Perhaps every state should be divided by the same, open source, simple and fair algorithm that creates deterministic centroids of people of close to the same population to maximize political competition, not minimize it.
This is something too important to be left to political hacks of some states but not others who seek to cheat at elections to favor incumbents.
This is spot on and cuts both ways. Much of the Japanese market's recent "performance" in US media is actually just yen weakness against the dollar. Strip out currency effects and the story looks very different. Same with European markets "performance" - we're often seeing monetary policy divergence rather than genuine outperformance in foreign markets.
Always check both local currency and USD returns when evaluating international markets.
MSCI Europe is up about 8-9% in euros so far this year—roughly the same as the S&P 500 in dollars.
But the euro itself has climbed ~10% YTD vs the dollar (≈ $1.02 → $1.12-1.18). So you get an ~18% gain if you invest in MSCI Europe in dollars.
Europe hasn't "beaten" US stocks because its companies suddenly out-executed; most of the gap is the stronger euro.
Not that it matters who’s "winning." My gripe is with US headlines that shout "Japan stocks are on fire" or "Europe stocks are on fire," when what’s really happening is that global markets are rising together and currency swings make one region look better than another.
Not in Australia. Our dollar has taken a beating too. I guess digging holes and selling property to each other at ever higher prices isn’t that interesting to the rest of the economic world.
Australia printed a lot more money relatively than the US from COVID-19 until now, largely to capitalize on a booming commodities sector. A factor that led to some do weakness.
But I think any weakness is temporary. With a stable government and abundant natural resources that will be even more sought-after in an AI-driven world and largely insulated from automation Australia’s long-term prospects look strong.
> Athletes consume 50g of fructose (and 50g of glucose) per hour
...while actively competing in endurance events. They don't do this while sitting on the couch in the evening. And it's very specifically only endurance trained athletes who do this, sprinters would never need to consume that much sugar.
It's a valid point though, a sugary snack or sports drink directly before hitting the gym for a hard workout session is probably fine, as long as your blood wasn't already saturated with glucose from earlier sugary foods, and as long as you don't have insulin resistance.
Sure, but that all has nothing to do with digestion, that has to do with blood sugar levels and what your body does with the energy afterwards. Digesting 30g of fructose per day is not any kind of biological limit, in contrast to what GGP claimed.
It's an incentives issue, not a community input issue.
If the only value you are able to hold onto is the development you do yourself, then you aren't going to block your neighbours from building up their property.
Members of the community have incentives to prevent high speed rail and apartment towers being built near them. The general population has incentives to build high speed and apartment towers in those neighborhoods.
But, because of community input, the voice of local busybodies is louder than that of the general population. It is a a disservice to democracy and prevents our governments from functioning in a way that serves the people.
Incentives have nothing to do with it. Until the community recognizes that other people have rights and that rights are rights precisely because you need no "community input" (which is an euphemism for "political interference") in order to exercise them, the results will be the same.
While developing a piece of real estate can cause a good deal of damage to the character of a community, ideas are much more dangerous in that regard.
Why is it then, that we allow for unscrupulous capitalists to disseminate ideas freely? Why isn't there "community input" into the things newspapers are allowed to publish?
It wouldn't actually. The contact in his phone (incorrectly added by Apple AI from a forwarded email) would be the same regardless which app he was using.
Instead, Signal (and this forked version) would have to do its own independent contact management, maybe based on in-person scanning of QR codes plus web-of-trust.
If only it would a- not ask you to access your contacts and b- accept when you say no instead of saying "we'll ask again later" (and then, indeed, asking again later).
Some people can do macro tracking, others don't have the discipline for it. For those people, shifting to foods with a higher satiety to calorie ratio is a better strategy, or intermittent fasting, or cutting out a food group. As you said, what matters is CICO, not how you achieve it.
I like to compare it to programming. If you tell a C++ developer that their software should have good uptime, your advice here is the equivalent of saying "don't have memory leaks, null pointer dereferces or use-after-free". Yes, all true, but everyone know that. What we need are behaviour patterns like RAII, an extensive test suite, running those tests in valgrind/ASAN, etc. that actually help in a forward looking perspective achieve this goal of not making those mistakes which lead to poor performance.
> For those people, shifting to foods with a higher satiety to calorie ratio is a better strategy, or intermittent fasting, or cutting out a food group.
If they can't have the discipline for CICO, why would you give them benefit of the doubt they have the discipline to "follow" the other methods? It makes no sense.
The discipline needed to resist hunger and the discipline needed to stick to a healthy and varied diet is wildly different. The second may require more thought and skill but a lot less willpower.
