If you watch the video at the bottom of the article, you'll see photos of several prototype FarmBots that do, in fact, exist.

I think there's definitely a market for this. For example, I'd like to have a nice vegetable garden in my back yard, but I don't have the expertise or the free time to do the work necessary to keep it healthy and happy. If I could buy a FarmBot at the local Home Depot, set it up, press "Go", and not worry about it until harvest time, that would be a pretty tempting prospect. And once the technology got cheap enough and reliable enough for most people to afford and install, anyone with some land could easily grow their own organic produce, exactly to their own specifications. For people who don't have their own land, neighborhoods could do slightly larger-scale versions of the same thing in the community gardens. Peoples' ability to feed themselves (rather than rely on buying food from large corporations) would increase, which can only be a good thing.

Longer than that. Fusion power has been hyped since the 1950s. From the article:

Heard that one before.

Fusion power has some real problems. After half a century of trying, nobody has a long-running sustained fusion reactor, even an experimental one. The whole "inertial fusion" thing turned out to be a cover for bomb research. There's a lot of skepticism about whether ITER will do anything useful. It's not clear that a fusion reactor will be cost-effective even with a near-zero fuel cost. (Fission reactors already have that problem.) It's really frustrating.

Fusion reactors are a pain to engineer. They have a big vacuum chamber with high-energy particles reacting inside, and huge cryogenic magnets outside. This is far more complicated than a fission reactor, and is why the cost of ITER keeps going up.

There is no "FarmBot". There is only a Kickstarter project to start a wiki to create a social network for talking about farming-related subjects, parhaps including talking about a FarmBot.

The basic office-type products for Linux still kind of suck. I've been using them since the StarOffice/SunOffice days, and now use LibreOffice. They've improved a lot, but they're still flakier than they should be, a decade after initial release. Nobody wants to fix the hard-to-fix, boring bugs which damage usability.

Oracle buying the remnants of Sun didn't help.

I doubt it. Any company that wants to stay in business will instead concentrate on making sure the car does not get into a position like that in the first place -- because once the car is in a "no-win" situation like that, it doesn't really matter what choice it makes, the company is going to be hit with a big lawsuit either way.

Before "Jaws", there wasn't much of a market for shark meat. Then demand picked up. Now, the shark population has dropped so much that sharks are facing extinction.

Will this be fixed on the next Patch Tuesday? I haven't been using the Windows macines much lately, but one is powered up, idle, and accepting updates. Will it fix itself?

Not to mention only in a society where energy was so plentiful that they could afford to waste 99% of it by sending it in all possible directions all the time.

Apparently they are retarded in Germany...

As usual, we have to go through two levels of blogs to get to the actual ICANN document. Which document you may find incomprehensible even if you know how DNS works.

This is in theory true, but there are proposed proof-of-work algorithms for which specialized hardware doesn't have a huge edge over general-purpose CPUs. Such algorithms require more memory than the existing hashes, and are designed to be highly sequential, so they don't parallelize easily. At least one altcoin claims to have such an algorithm.

Any algorithm that requires a significant amount of 64-bit floating point computation and lots of memory, like a big matrix inversion, would be reasonably ASIC-proof, simply because that's a task CPUs are designed to do fast. An ASIC that could invert big matrices would need superscalar FPUs, which makes it as complex and expensive as a CPU with comparable performance.

So far, nobody seems to have devised a "minable" algorithm based on matrix inversion, but that's a place to look for one.

communication is not language. Language requires the ability to be infinitely expanded.

True, but not particularly helpful since 99% of the time the user does not know what he's doing (at least, not from a computer-security standpoint -- all the user typically knows is that he's trying to accomplish task X, and here's a dialog that says it can help with that task if he clicks OK...).

Who knows? The whole point of a zero-day exploit is that it takes advantage of a previously-undiscovered flaw. So there is a bug in your email reader that causes it (under certain circumstances) to automatically activate malware, you probably wouldn't know about it until after the fact -- and if the infecting software was subtle (hi NSA!), probably not even then.