xbkingx

Blu-rays can take up 25gb each, so just a decent collection of those could easily consume most of one of these drives. If you want to do basic model tuning in stable diffusion, each model variation can take 7gb. This level of storage would mean you could almost setup a versioning system for those. And finally, any work with uncompressed data, which can just be easier in general, could benefit from it.

I would love that, too, but it's not possible today. Everyone would offer 25 year warranties, close up shop in 5 years, and reopen as a new subsidy or company.

The only way I see it working is to hold some large portion of the revenue in a trust and relinquish it to the company over the warrantied lifespan. The company would have to operate at a loss for a while to books those reserves, so there would have top be something like a zero interest government loan to cover the cost, which can't be escaped through bankruptcy.

Or maybe a contract like the shitty cell phone plans in the US. Buyer agrees to pay for the full price of the appliance over the warrantied lifetime in installments. If you want to sell it or trade it in early, you either have to finish off the payment or transfer the contract. The company would have to service the product (within reason), or the contract is voided, releasing the buyer from payment obligations. Again, this system can be easily gamed, too, in today's market, but I just can't imagine a scenario that doesn't require a major paradigm shift.

I do a decent amount of 3d printing and I cannot count the number of random letter Amazon brands for filament that have popped up over the last year. Most are simply rebranded waste from larger manufacturers. Once the product gets below 3.5 stars, the brand disappears and a suspiciously similar new brand pops up with the same spool design and 20 5-star reviews overnight.

lol I wish it was this easy. I got through the simulator in 6yrs 11mos on the second try. At no point was hope above 40%, except once early on (ended at 33%).

The funny thing is that I had 1 conf paper, 1 major result, and 1 figure left over. That's a good year extra, so I assume a perfect game would be to get the 3x papers and GTFO (which is the second best outcome, after not enrolling). There were a couple folks I knew that made it out in 5 years, but more that took 7+. Our lab was notorious for taking over 10, which I skirted by.

Like others said, this was lacking outside events (social/political junk). Hopefully version 2 will take into account: at least 1 family death and 1 additional tragedy, at least two months lost to helping or waiting for help from another grad student or post doc (they did have the lab equipment breaking, which was good to see, but missed the lobbying for every little purchase), at least one scope change, a half dozen favors to gain some political cache, a few experiments and/or rewrites to satisfy faculty members that just read about a technical issue they should have known, but didn't so they're highly sensitive to it, at least 6 months of arranging the data/results in a way that faculty can understand, 3 months of arguing that the lab standard procedure for some basic component is a decade out of date, a few months worth of preparing premature data for unnecessary meetings, one (and it better be just one) instance of an offer to help getting waaaay out of control, the hope boost after your first big conference and subsequent conference hope drops, the drops with each thesis defense from folks a year younger, etc. There's more, but that's off the top of my head. Oh, and that slight boost in hope when you hear someone else has a worse problem than your current one. That's a fun one.

Tip for those interviewing - ignore all the year 1-3 folks. 1 and 2 are basically undergrads plus some extra classes. 3 probably hasn't hit the first pile of bullshit yet. Find a year 5 or 6 in your field and talk to them alone. There's a reason they generally don't have senior grad students at recruiting events, and it isn't because they're too busy. Talk to them long enough to get to their exhausted attempts to rationalize some aspect of the experience. If their demeanor doesn't change, you might be safe. If they start hemming and hawing, that's a problem. They haven't even gotten to a specific, non-personal problem and they're having trouble keeping up the facade. The layers are: 1) Hey, social event, I get to take my mind off lab problems. 2) Getting a little boost by talking to someone still excited. 3) The quiet whisper, "Let me give you some advice." 4) The realization that there's nothing but lab to talk about. That's the threshold. 5) The rationalization alpha - The view from 30,000 feet isn't terrible. 6) The rationalization beta - The rundown of broad problems they're having. This is the point where they will probably, as if by magic, remember that thing they were going to do needs to be done now. (I've got some analysis running I need to check, I need to feed some lab animals, I promised my parents I would call, I told a lab mate I'd help them with this thing and will be up all night, etc.) 7) The rationalization gamma - Specific cases of major problems they're seen other have. 8) The rationalization delta - Specific problems they're having.

The moment I saw the term "SEO" it was like a stopwatch ticked on until search was dead. It used to be frowned upon to do little tricks, like keywords in a tiny, transparent font picked up by crawlers, but not seen by users.

When gaming search engines became a profession, the end of search appeared on the horizon. Guess we're headed back to web rings and link indexes (which will be consolidated, heavily monetized, and abandoned). If we're lucky, we'll be back to dialup BBSes by the 2040s.

And the people that put significant effort into their work will reupload it to other services, if they haven't already. Similarly, exceptional work is probably saved by random folks elsewhere.

What? You don't want to watch the full box set of Star Trek TNG between when you push a button and the animation ends? That ruins my plans of designing a scrollbar that plays 45 random tracks from you music library between lines scrolled.

OTOH, it IS very satisfying to realize you don't really need to do whatever requires those interactions in the first place.

It speaks to the more serious problem with vaccine hesitancy - the more virions (individual virus 'particles') that exist, the greater the chance of a mutation that finds a way around our defenses. It's just basic evolution.

The variants arise randomly and proliferate with the current major strains. The infected population adapts to reduce the transmission of the most virulent/contagious strains. Selective pressure (tug of war between infecting people and people fighting off the infection) increases on the new variants until one or a few maintain or exceed the transmission of the progenitor strain.

That's the problem with the, "I'll get over it/I'm not worried/My segment of the population doesn't die from it" mentality. The more infections - subclinical, asymptomatic, severe, fatal, undetected - the more rolls of the dice. We (humans) are selecting variants that are worse for us, hoping we can snuff out the infection before some key mutation that eludes our immune system and/or testing develops.

Two other thoughts with internal conflicts/points worth mentioning - First, recovered patients should be more resistant to new strains. Their immune systems threw everything at the virus to defeat it, so their response will be more diversified than those with mRNA vaccines targeting a specific protein sequence. (The magnitude and usefulness of the variations in immune response can negate that advantage.) Second, the reasons that a 'novel' virus is dangerous are that we, as a species, don't know if we can (naturally/innately), and we don't know how much the virus can change in protein sequence (to evade our defenses) or our response to infection.

Anyways, I'm rambling. Not a virologist, but a PhD (and as such, I think I know more about stuff than a really do) and that's how I think about it. :)

I kept waiting for it to get better than a light brown haze with flashes of green... but it never did. The fire looked okay, but even that looked weird at the beginning of the sequence.

What really made me laugh was the blue smoke followed by a headline implying that was all at night. The AI filled in pristine blue skies and inverted the wreckage colors. It actually made up history. I'm calling it: Unintentional automated disinformation.

I think it's more of an emergent pattern.

Mutant ABC is the target. First Mutant A becomes the predominant strain. There might be some Mutant AB or AC in the population, but the controlling factor is currently Mutant A.

Eventually, Mutant A therapies/vaccines are introduced and Mutant A starts to lose its hold. Mutant AB isn't treated by the new treatments (or they're less effective, or Mutant AB goes undetected longer, etc.). Mutant AB becomes king of the hill. Repeat with Mutant ABC.

All this would assume that there's no Mutant B or Mutant BC or Mutant AC that increases in prevalence. All the mutations would have to build up over time or occur simultaneously by coincidence (or you could get a combination).

It's just basic evolution - organism follows the niche. We have a hard time recognizing it because generation turnover is so rapid. Instead of taking ~30 years for 2.5 offspring, it's a few days for millions of offspring.

I don't think mode of reproduction matters much when we're talking at this scale. While the virus creates stable clones across generations, I would think they're more susceptible to mutations and any viable mutation is quickly amplified exponentially. Just like the old high school bio experiment to demonstrate antibiotic resistance, only your respiratory mucus membranes are the medium. Those bacteria reproduced asexually.

I had the same problem with and older Netgear router. It turned out to be a bug in the scheduling settings. It had the ability to disable network access between specific times on a daily schedule. The feature was enabled by default, but no times were set to disconnect. You'd think that would mean the connection would never go down, but a firmware update changed that. Every morning at ~3:00am the network would go down for 1 minute. Disabling the feature entirely resolved the problem.

Looks nice, but no ability to set default playback speed means I'll stick with (the rarely updated) Vanced. I never watch YT content at <1.25x speed unless it's to hear a music track.

It also looks like it doesn't have the option to login and sync, so I'd be seeing already watched stuff on between desktop and mobile. I understand why some people wouldn't want that, but I use the feature all the time and would like the option.

I might try it as a kind of backup app for things I want to track, but don't want suggestions for (mostly technical videos).

The problem for me is that I honestly can't think of more than a half dozen instances in the last 20+ years where ads made me aware of something new that I actually wanted, and zero of those were web-based. I can't think of a single time ads swayed my opinion to purchase something I didn't want or caused me to buy something earlier than I would have otherwise. At BEST there have been a few ads that have led me explore options within that market, but the company posting the ad had zero impact on my decision (it almost deters me, since that ad money could have been spent making the product better or cheaper, so the competition is probably a better buy).

Why would I want even 1% of the content I'm consuming to be polluted with irrelevant garbage? That's a huge amount of time and mental effort filtering out stuff like toothpaste ads because I googled directions to my dentist's office.

Nope. Ad blockers on EVERYTHING and custom CSS for sites that whine about ad blockers or display a sea of whitespace to inflate page view times.

I had a side job as a tech reviewer ~15 years ago and I can tell you why this is never going to happen. It's a classic case of "works in theory, but impossible in practice." What you're actually asking for is a database of every permutation of hardware, hardware revision, firmware/BIOS, benchmarking software, benchmarking software version, driver version, OS, OS patch, user selectable feature state, etc. is tracked and tested. It's a nearly infinite data set size, and you'll still miss things like ambient room temperature/humidity for CPU thottling, that one finicky USB device you spilled water on that sometimes disconnects randomly, and specific workflow quirks.

"Objective" benchmarks were an almost tractable problem 15-20 years ago, but with the way modern OSes run background tasks, access network resources, and perform self-maintenance, it's even more difficult to bridge the synthetic-to-real world divide.

So, you select a point of reference (e.g. - new GPU), you choose the most common system components and a selection of components that tell a story (usually "if you're building a new PC now, here are the options"), you assemble the current versions of all software/drivers, and run your tests x times. The good sites will have "sub-stories," like specific workflows or new/changed features, but once you get to around 4 of these, you start hitting too many permutations to clearly communicate the significance of your choices.

Even if you crowdsourced it, it's A) extremely difficult to verify the integrity of results and prevent manipulation from marketing departments or brigading, and B) an unrewarding, tedious process where the best practice is to leave the machine untouched for hours. Most of the crowdsourced benchmarking sites are set up to be little competitions and sanity checks when overclocking. For a pure review of shipping hardware/software, you run the benchmark and you're done. It isn't particularly fun, requires invasive cataloguing of system specs, and isn't very new-user/first time builder friendly.

Dynamic graphs would be nice (I love them), but I suspect many review sites have run the numbers and found they reduce overall web metrics. Many (I would argue most) people don't notice interactive page components, aren't interested enough to turn their quick article skim into a deep dive, or are reading in less optimal settings (e.g. - on phone on the toilet). Instead of getting 7 page views for 1 minute each with separate graphs, you're getting 1 page view for either 1 minute (probably 70%) or 10 minutes (probably 30%) with a dynamic graph.

I'm a total data junkie and I WISH there was a good solution to these problems, but there isn't a practical implementation that I've seen or dreamt up that doesn't have enough variance to render the point of having such fine-grained data moot.