Because the local environment is going to get more and more cuthroat and backstabby as people feel pressured.
1,222 karma · joined September 5, 2019
Well, the dynamic they've introduced most strongly incentivizes competent [read: employable] employees to leave so certainly that will improve overall team performance and they won't need to continue layoffs /s. To call it a death spiral would be excessive, but negative-divergence quality spiral just lacks that special something...
1) A cool research language packaged into a VS Code plugin 2) A cool research compiler for Python + numpy 3) A cool generic ISA project
will kill C++? Maybe for the author's _very_ specific use case of blindly copying things from SymPy, but really I think you'd have better odds betting on a D resurgence.
No. Here's a protip which applies to every country across the globe: if you're donating money to people doing things your government decided is illegal with the express purpose of enabling continued illegal acts, you're gonna have a bad time.
For actual production design like the OOP is talking about, you would likely go with whichever option is cheapest overall (which in my experience has universally been the standard green).
The threaded fittings however are a different story. Assuming you didn't use stainless steel on stainless steel fittings [0]:
* You can find the leaks by spraying soapy water on suspect connections and looking for bubbles (wash off and dry with clean water afterwards unless you want some serious corrosion). You can also buy a jar of noncorrosive propylene glycol based leak check fluid at your local hardware store for somewhere around $10, which is a strictly better option although it's a bit harder to clean off.
* Undo the connection and throw away every fitting you can replace (you generally should not reuse NPT threaded connections unless you know what you're doing). Clean all fittings until they look brand new and with no visible debris on/in the threads.
* Watch some videos [1] and remake the connection using the proper amount of tape and sealant and appropriate torque [2]. Only tighten the fitting. If you loosen it even a bit during the process, undo it completely. Clean both sides, reapply tape and dope, and try again. Let the sealant set ~24h and retest.
[0]: Stainless steel pipe connections are a special case because they tend to cold weld before they're fully tightened. There are ways to mitigate that but the short answer is don't use them if you don't already know how.
[1]: https://www.youtube.com/watch?v=whw9ApDJpJo
[2]: NPT connections really shouldn't be torqued (you instead use turns from hand-tight for properly cut threads) but if you're having trouble find a torque chart for your fitting material (copper, brass, steel, etc) and follow that.
As far as power density goes, the i9-13900K seems to be 335W / (257 mm^2 die area * ~1mm water thickness) ≈ 1.3e9 W/m^3. The power density of the sun in comparison is around 2.8e2 W/m^3, roughly what you'd find at the center of a large pile of compost (although the temperature at the sun's core may be just a wee bit higher).
Instead he called Putin to talk about "peace" (read: appeasement) and then posts literal Kremlin talking points to Twitter and only mentioned getting money to maintain starlink after getting a ton of people telling him to fuck off.
A charitable interpretation of his actions is that he's occasionally unable to contain his narcissistic traits, a less charitable interpretation is that he's looking for leverage to get his dick out of the hornet's nest that is the Twitter deal, and the least charitable option is that Russia is applying some leverage to force his hand. No matter the reason there's a good chance he'll end up regretting that (to the extent he's able to feel things like regret anyway).
With DALL-E mini, the classic opening sentence of the dark tower "the man in black fled across the desert, and the gunslinger followed" produces evocative (if abstract) artwork. With DALL-E 2? You get a message that your prompt is inappropriate and your account will be reviewed for termination if you continue. So they don't want guns, okay whatever.
But it eventually impacts pretty much any concept you want to try. I was exploring scifi art and wanted to see what it would do if I tried to get a stylistic fusion of old-school soviet spacecraft aesthetics (exposed structures with bulbous pressure vessels housing the controls [among other things]) with western equivalents. Fusing disparate design philosophies in a way that actually feels creative is a task at which DALL-E 2 occasionally performs superbly and I was really looking forward to playing with the concept -- dreaming through the machine of a long-lost potential future.
Nope! Turns out "Soviet", "USSR", and IIRC now "Russia" are straight up banned words. I burned up quite a few prompts before figuring out that was the trigger.
And then there's the cost. During the closed beta they gave us what's now several hundred dollars a month in access. That translated to less than an hour a day of exploring DALL-E. It felt inadequate when it was free, and now they're asking for hundreds of dollars a month for it? No.
The pricing model makes sense if you're trying to generate bland corporate artwork for some random webpage but as a creative tool it just ensures that the only people really engaging with it have substantial financial backing for producing exclusively milquetoast ``artwork.''
I can't really comment on that as 1) I only skimmed a couple sections of the paper and 2) my field is more applied math than pure. I'm not attacking or dismissing anything, just asking how the claims made are substantiated.
> However, since it's able to model experiments in the Large Hadron Collider, I would assume that it makes predictions about the positions and velocities of high energy particles.
Do those models cover experiments which have yet to be run, and more importantly are they predictions which differ from those of the standard model?
There are a lot of neat theories which model experimental work we already have results for, the problem is using said theories to make falsifiable predictions which are both realistic in terms of actually finding a way to do experimental verification and are different from predictions made by the standard model. Nobody really loves the state of physics as-is, but moving on to something else requires meeting those two conditions which has been an insurmountable hurdle as of yet.
Unless we end up with refinery bottlenecks (like we more or less have now) or political incentives (although the odds of that seem alarmingly low at this point).
> The consumption would be lower so price would be too.
Is this expected to hold as oil usage drops over the next century?
If certain justifications are motivated to prosecute this as some form of murder (which is certainly within the realm of possibility now), Google's location history having holes in it when other data sources indicate that someone's phone traveled to an abortion clinic may not work in their favor. They might be better off perturbing the precise location to something else nearby.
This is basically trying to solve the fugitive state act with technology. Nice idea, but seems to seriously underestimate its adversaries
> Speaking of naïve realism, what evidence
You can really only prove this type of thing in a single direction, by providing evidence that there was fraud on such a scale as to have swung the results. Otherwise, one could endlessly claim that the election thieves just covered their tracks so well that you don't have any evidence of your assertions regarding their behavior.
The burden of proof was on the ones making accusations of election stealing, and as was reflected in how they lost case after case regarding this, _they did not provide it_.
0: Take a guess as to what those look like inside.