You only need one piece of (timeless) advice regarding what to look for, really: if it looks too good to be true, it almost certainly is. Caveat emptor.
Seriously, ending up with nothing is always a risk you run when buying something advertised as non-functional in the hope of fixing it or recovering any undamaged parts. The fact that the components on this card weren't original is almost irrelevant, because the result would have been the same if they were authentic but damaged beyond recovery.
I suspect they're making an unwarranted assumption that the experimental patient ended up with high cholesterol due to excessive consumption of animal products (rather than, say, a genetic defect that would cause them to overproduce it regardless of diet) and applying some typical vegan arguments regarding livestock farming. No need to listen to them.
There are still a few brands of dumb consumer TVs on the market, although they're becoming harder to find. Ars Technica did a roundup in December.
Even their older, simpler fridges are crappy. We bought one because our previous fridge conked out in mid-pandemic when the selection of new appliances was low. It lasted about three years before developing an issue that would have cost us more to fix than just replacing the damned thing. So we replaced it with some cheaper probably-Chinese brand I'd never heard of before and will never buy another Samsung appliance again if we can help it. AI will just add expensive, useless functions on top of their already poor design and dubious manufacturing.
In other words, if these become the only fridges in existence, I may just try to find out where I can purchase an old-fashioned icebox.
According to records obtained by the group, “it’s often impossible to tell which parts of a police report were generated by AI and which parts were written by an officer.”
This does not give me a great impression of the literacy level of American police officers. Another good reason to stay out of that country.
If you have no internet but want your music as a file, that's how you go about it.
They picked the wrong history, in my not so humble opinion. The AI situation looks more like the dot-com bubble, recycled.
Lacking government regulation in the largest markets, proper separation will never be enforced, because it isn't to the manufacturers' benefits. And that probably isn't going to happen until hacked infotainment systems kill enough people to draw attention, unfortunately.
Animals, including humans, have sensors for pain (nerve endings), and a series of routines in our brains to process the sensory data and treat it as an unpleasant stimulus. These are not optional systems, but innate ones.
Machines not only lack the required sensor systems and processing routines, they can't even interpret a stimulus as unpleasant. They can't feel pain. If you need proof of that, hit a computer with a sledgehammer. I guarantee it won't complain, or even notice before you damage it beyond functioning.
(They can, of course, make us feel pain. I just spent the last hour trying to get a udev rule to work . . .)
For some people, recreating the form factor of a book is the point, regardless of its convenience or cost. I'm sure whoever put this thing together was quite aware of how mainstream e-readers are built and didn't want that, or they would have bought a Kindle or a Kobo.
Attempting to defeat browser fingerprinting (you can never be 100% sure you've defeated everything) without TOR is kind of an advanced subject, yeah, and one of which I have only shallow knowledge. A lot of it is Javascript-dependent, so allowing Javascript only on a whitelist basis should help (but is too tedious for a lot of people). Deliberately pissing in the pool by varying prominent identifiers like the User-Agent string should help. Canvas poisoning. Specialist browser extensions, some of which may be more effective than others. Running the blandest default-settings browser possible in the blandest possible default-settings environment (a container or live media inside a VM) could conceivably cause you to vanish into the noise, but may be highly inconvenient.
It's worth considering who is likely to be interested in going to the trouble of browser fingerprinting in the first place. Small players have little use for the information and aren't likely to accumulate enough to sell it for much money. So the problems are going to come from ad networks, large digital networks like Google and Meta/Facebook, possibly CDNs and service providers like Shopify if they think it's worth their while, maybe some governments, and completely dishonest scam sites that think any money is good. Some of these can be avoided altogether if you work at it.
Someone too lazy to update their listings to reflect a rising sticker price (or not wishing to do so for other reasons) isn't too good to be true. If they're an established business selling new-in-box items at more than the wholesale price they would have paid (around 50% of the lowest sticker price the good's ever been sold at isn't a bad estimate), then you may have found a genuinely good deal.
It's when someone starts selling at below their cost (unless it's obviously to clear out old inventory or the like) that things get suspicious.