HakFoo @ HakFoo @lemmy.sdf.org

Posts

10

Comments

487

Joined

2 yr. ago

Stop selling it a loss.

When each ugly picture costs $1.75, and every needless summary or expansion costs 59 cents, nobody's going to want it.

You might look into the Kailh Box switches. The click leaf adds a distinct tactile bump, plus the sound factor. The Box Navy is almost painfully tactile.

I wonder if it's just so far outside the scope of what capitalists know to risk model that it's not a compelling bet. The "it's 20-50 years on" timelines are also a hard sell.

If you win by delivering a reactor, you unlock so many knock-on effects.

Does your business immediately get subsumed on national security grounds? Does a massive reshape of the energy market cut off your current gravy trains? How do you monetize "too cheap to meter"?

Researchers don't care. They're in it for the cool project.

Try RiscOS for a glimpse of a world most of us missed.

The worry is that it feels like we're moving past a consumer-directed economy and not in the wholesome Soviet five-year-plan way.

The almighty market has figured out how to collude and cram stuff that people don't want down our throats (what consumer wanted everything to turn into an enshittified subscription?)

Real people may not want AI slop, but if there's enough of a sense it will make line go up, we're getting slop.

On the other hand, this factor might be the salvation: the current AI market is full of 2000-era-dot-com business models based on selling at a loss and making it up on the promise of global domination later. If the VC money dries up, and every "delve" costs whatever the actual amount it costs to drain the oceans and oilfields to pump into an array of Quadros plus sufficiently reimbuse all the ghouls that bankrolled the project, maybe the line doesn't go up anymore.

The Very Busy Tactical.

I suspect it could be seen as a proper noun.

If Acme and FooCorp create a bridge between their private network spaces, it's an internet (common noun) but not the Internet (proper noun, referring to the one with Goatse).

Let's find an English teacher. And yell at them for forcing us to read the same terrible novel in both 10th and 12th grades. Maybe after that, return to this subject.

I'd say it's a bad thing because it's the wrong threat model as a default.

More home users are in scenarios like "I spilled a can of Diet Sprite into my laptop, can someone yank the SSD and recover my cat pictures" than "Someone stole my laptop and has physical access to state secrets that Hegseth has yet to blurt on Twitch chat". Encryption makes the first scenario a lot harder to easily recover from, and people with explicit high security needs should opt into it or have organization-managed configs.

Even the Grinch didn't go on TV to tell peoole he was stealing Christmas.

I thought the appeal of fentanyl was that it was so cheap to produce that you didn't need to cut it with other drugs.

To be honest, I'm amazed it took til Biden before we saw more pressure on TSMC as a flashpoint.

Even if we're on nominal good terms because they're a capitalist democracy, nobody likes single points of failure (earthquakes and industrial disasters happen even without geopolitical tensions)

But we've handled it miserably-- throw some money at Intel who can't innovate out of their gilded cage anymore, and try to get a few TSMC facilities stateside-- when we should have been trying to completely diversify the supply chains with new players and new geographies.

In fact, it's amazing that we lost the concept of second sourcing. That ensured no one vendor held you hostage. Like 8 different firms made 8088s, on up to the 486, but after that it dried up fast. You saw a few IBM badged Cyrix 6x86s, but who else sells a pin-compatible Ryzen?

I hope once China gets far enough up on the tech curve, they see distributing fab tech as a BRI programme. No reason your next bag of 74LS04s, or the 30-cent MCU in your thermostat, can't be made on a 28nm fab in Burkina Faso.

Then you have the NexGen Nx586, which is arguably 386-like in having no FPU, but ended up being the ancestor of most modern x86 CPUs by decomposing complex operations into RISC-esque micro-ops.

The Global Foundries split was probably a way to get AMD out of the hyper-capital-intensive fab business. And without a tier-1 customer, Global had less reason to pursue smaller nodes.

Intel has that national-champion thing to keep it afloat. I can imagine there are defence contracts that will never go to a "TSMC Arizona Division" and they'll pay whatever it takes to keep that going.

Once an egg sac hatched at our communal postbox and it was magical, all these tiny yellow spot sprite things bumbling around and setting off. Put a good narrator on and people would have watched for hours.

I think it's more the £50 notes. Much like using a USD100 note in the States, it's a bit big for most daily purchases.

I ended up dumping most of mine on a couple expensive souvenirs in shops expensive enough that they'd deal with it or breaking them in banks.

Some people like to use a stored balance as a financial discipline tool. Don't put a "real" funding source on the account and then you can only spend the $100 you committed to, and not go whale-mad and drop $500 on premium currencies.

It's easier just to price in the fee than having to shut down or retool a project.

The problem with attribution is the difficulty of 1000% accurate compliance.

If you grab 100 lines of code from a repository, or five paragraphs from a story, there's probably a claim there. If you grab a single word, there's probably not. But in the middle, there's a paralysis of uncertainty-- is n lines similar enough to create liability? Can you remember where you saw what reliably? You end up with a bias towards "over-attribution" and it becomes difficult to pare it back. Does everything need a full Git-style commit history? Are we forever stuck keeping a credit on a project because it's difficult to prove you've fully scrubbed their contributions?

Focus on how we pay artists (ideally lush grants) and forget about credit. Maybe establish a culture where it's voluntary and acceptable-- that people feel that they're allowed to cite their raw materials, and reuse doesn't make the work lesser-- but don't try to use the courts to force people to try to remember and track where they saw something when they just want to create, or it creates a hostile environment.

Hey, the broken clock's right!

IP law always had a built-in scale pronlem. Without a registration-required copyright model, and probably some sort of mandatory licensing rate system, the sheer logistics of finding and arranging rights made a lot of business models inpractical. (For example, why aren't modern bookstores just print-on-demand kiosks, or streaming services have All The Content? In large part because it would cost thousands to track down owners and negotiate terms for $1.87 in royalties multiplied by every item in the catalog.)

This was ignorable for a long time, or even a commercial advantage for firms with access to large, pre-negotiated catalogs. The AI boom created a surprise market of non-incumbents who need to get access to a lot of IP in a streamlined manner.

If we open the door for bulk IP clearance to grant the AI bubble a stro ger legal footing, it can also allow other, potentially more interesting business ideas to slip through.

It's impressive how much of the recent deportation scheme leans on the cooperation of one tin-pot state.

There aren't that many other places with concentration camps conveninently located and leadership ready to deal. It's unlikely they'd build them domestically, it would take time, cost a fortune, and not achieve the explicit "we removed the evil foreigners" goal.

It would be interesting to see what happened if someone said "we'll pay you more than what America is paying to close the door." Would he have to just knock on every presidential palace in the hemisphere looking for a new partner? Try to scale Guantanamo 100x overnight?