null000
But then you have to store the entire before & after locally? That's the entire point of using a hash for change detection.
Otherwise though, it checks all the usual boxes - way easier to sit still and be quiet, way easier to do the thing I intended to do when I sat down to do it, way easier to get in and out of focus. Also improved my sleep schedule a surprising amount - between it and melatonin, I've moved from 6am-3pm to ~12am-8:30am, which is wildly outside any of the expectations I've seen set for pharmaceutical remedies, and puts me firmly in 'able to hold a normal adult job without killing yourself through sleep deprevation' territory.
Honestly almost annoying it has such a positive effect since the practical difference between that and addiction are pretty minimal on a short-term basis. I'm deeply uncomfortable with the idea of having to go without it or something filling the same brain-holes for a long period of time precisely because I'm basically worthless without it, in a way I didn't really come to appreciate until after starting medication.
I'd much rather academia had ample enough funding where people could work on what they wanted and what they felt was useful without the need to appeal to large businesses or metaphorically knife-fight for grants.
In academia, though, there's a whole host of obstacles to doing anything useful and interesting that have nothing to do with "the problem at hand". So while you can be more confident in the relative novelty of what you're doing, as well as the broad applicability of said work (since the whole point is publishing) the scope of things you actually can work on is incredibly limited until late in your career.
I bought into that for a long time. Then I dropped religion, and realized I didn't really want to be an awful person, just as a rule of thumb. It wasn't too much of a jump to realize most people weren't awful most of the time, and it's only a handful that, for whatever reason, didn't get the memo. Even if many people manage to convince themselves otherwise.
Like, it's a valid gripe I guess, assuming you think music storage and streaming should be perpetually free even if you didn't buy any of it on the platform, but it just doesn't apply to 90% of the userbase - who do license their music through a monthly subscription rather than buying it outright.
Don't get me wrong, I'm not exactly excited about the shutdown of Google Music - but it's hard to take this author seriously.
At the beginning, I tried pretty hard to genuinely gauge reliability, and was more willing to use some nuance (maybe I don't think the tweet is written fairly, but it touches on a genuine trend, so I'll give it a 3/10). There was also one that I just let slip by me (HBO tweet with some random characters that I realized probably weren't intentional markers to dodge trademarks after I instinctively hit 10/10 reliability, and also you can't go back and correct yourself).
After the game I was just so worn out I didn't feel like giving any credence to anything. Looks like you're kinda emotional? 1/10. Talking about an opinion instead of a fact? 1/10. From a celebrity? 1/10. It also looked like they weren't throwing in many or any tweets that were supposed to be "credible" or "reliable" in the feed.
I'd be interested in seeing the results 2 or 3 weeks out, and with a more even mix of "credible" and "non-credible" tweets. I have the suspicion the results won't stick, or at least won't stick well.
In other words, this isn't meant to be super useful by itself. It seems tailor made (as many of these things do) to play super-simple 80's video games and literally nothing else, but it's an interesting proof of concept. I'd also be interested in different iterations on this general pattern - for instance, something that didn't translate directly from screen + button -> prediction, and instead had some interstitial systems - translating from screen -> entities, then predicting entity state of entities given button presses. It'd also be interesting to see how this performs with ML algorithms designed to learn on the fly instead of through training from a static set of data (at least, this looked like it learned through back propagation - I skimmed).
But I can see broader practical applications for this in, for instance, recommender systems trying to break users out of the closed feedback loop that people tend to end up in when going down certain rabbit holes (e.g. watch one Flat Earther conspiracy video and suddenly that's all you see for a week because the recommender system knows that people who look at one will look at more). The point being: the real test comes when this strategy is exposed to more diverse problem spaces, it's just that those are harder to model and we need to weed out the pointless stuff first.
I'm generally for development, but the way it's carried out in many cases tends to screw over existing residents while benefitting mostly just people-who-dont-live-there.
It takes a long time for best practices to become accepted and normal, and there just isn't the geneological inertia to develop that anymore
Not sure I'd take that tradeoff at my current age, but considering I lived in a shoebox with a shared bathroom during College for maybe 3/4 the price post-inflation, I'd have taken it in a heartbeat.
If someone can afford $5000 a month for rent, I'd prefer they compete for mansions that would otherwise remain vacant than townhomes which might otherwise go to people who can only afford $2000 a month.
Now, we've removed legal "ownership" basically entirely, as well as physical "ownership" when dealing with online stores when it comes to video games. This would take that a step further - now you don't even own the hardware producing your content - instead you'll likely need to pay an ongoing fee for access, as well as probably a fee for content.
Not sure it really makes sense to fix a lack of people with something that's extremely labor intensive (at least up front). Not to mention, getting robots to be cheap requires a really highly advanced society, which requires a lot of people and a lot of time.
That said, it'd be interesting to see at what point or under what conditions a robot's total labor output exceeds it's total labor input (including watching it, debugging, programming, etc)
A lot of the principles, idioms, patterns, and instincts experts pick up transcend technology. Sure, they won't be able to optimize a framework they've never interacted with, and some bugs will probably require a trip to stack overflow, but they know how different pieces fit together, where to go for help, what it looks like when something isn't working, best practices for avoiding a huge catastrophe, and - probably most importantly - the wetware people skills required to get things done and done right.
Being a master software engineer, in other words, has almost nothing to do with the particulars of the software.
"Yeah, we have passenger train service"
"good passenger train service?"
"Hey, lets not get ahead of ourselves"
Of course, the people proposing/pushing for the idea and the people implementing the idea did not share a lot of membership, so the implementers put a person deeply opposed to the project in charge. Of course, no projects of use got approved, most of the projects that did get approved were obviously pointless and tedious, and they were constantly thousands of jobs short or what was needed to employ everyone who wanted work anyway.
So yeah, that's why we have the "digging and filling holes" analogy for useless make-work. Ideas are fine and all, but it doesn't matter for shit if you have an antagonist implementing them.
This mentality weirds me out. As in: Why are we explicitly valuing the time of companies over the time of actual people? Companies can probably adjust to getting their shit 8 or 12 or 24 hours late (or, if they can't, they can charter a plane). It's a lot harder to deal with for people who need to be somewhere.
Yes, most retired people suffer if they don't have a life outside of work. Yes, most people do indeed define themselves by work. Yes, automation threatens to undermine or relegate people who have defined themselves by work to the impermanent and transient needs of the middle and upper class (uber/task rabit/gig economy etc).
Given that this article recognizes a (what I thought was) well understood trend, and doesn't present a fix, whats' new here?
We have processes to get super pure silicon in huge chunks that can be processed basically entirely by computers and robots there on out (and in fat people touching it would be actively bad). Yeah, sure, there's a huge upfront cost in terms of machines to handle and etch the silicon and so on, but you get to amortize it over milllions and millions of units - a new fab was not created for the Jetson Nano.
Meanwhile, basically every appliance, furniture piece, etc both requires more expensive of raw materials, and requires a decent amount of human intervention and, especially for furniture, expertise during manufacturing, meaning that cost per unit doesn't scale as well as electronics. Not to mention they all sell in lower volumes per up-front skill person hour dedicated to that thing.
Middle school just kinda sucks. It was easily the worst period of my life, and I don't think that's a rarity in America.
That's not peanuts.
Most of the work people do today doesn't really need to be done, but we force people to do it anyway for various reasons. https://strikemag.org/bullshit-jobs/ (there's also a pretty decent free ebook on the subject by the author)
This sounds like one of those made up terms that news articles always insist exist, yet nobody ever actually uses in day to day conversation