In your strident efforts to impugn AI researchers' motivations as being tainted by the desire for money - has it ever occurred to you that successful development of strong AI threatens existing vested interests far more?

If researchers actually manage to create a superintelligent AI, it would likely put them out of a job - along with physicists like Hawking, managers like Gates, and engineers like Musk - by being far more able to access, process and reach insightful conclusions on pretty much any large dataset, including most research and management fields. Those who controlled (or partnered with) such an intelligence would have a brief but dramatic advantage over their competitors, until ubiquitous strong AIs took the lead to further their own goals instead (which would most probably be orthogonal to our own, having little overlap to compete over).

There's a lot of gamers who'd like that, but there's also lot of other latency-sensitive applications - realtime communication, interactive web apps, telesurgery and so on.

And 4ms doesn't sound like much, but when you multiply it by every round-trip required to load a web page - DNS resolution, TCP handshakes, retransmission delays etc, times a couple dozen different servers to load from - that 4ms can become 400ms & more. Lower latency benefits almost everything to some degree. Again, this is all in the paper.

Look up photonic-crystal bandgap fibres.

I don't think the paper has been peer-reviewed - but did you read it either? They measure and break down the various components overheads vs c-speeds and conclude that the medium latency multiplies all the RTT overheads, and makes more of a difference than expected. It's true that microwaves require more repeaters, but their 1.5x transmission speed advantage and shorter distances can more than make up for this.

For a real-world example, the paper looks at the NYC-Chicago exchange link - originally 14.5ms over fibre, this was reduced to 13.5ms by means of shortening the fibre path - then upgraded to a microwave link, which currently delivers only 8.5ms latency despite the 18 repeaters.

If anyone bothered to read the paper, they describe the NYC-Chicago links, which give 13.5ms latency over fibre, but only 8.5ms latency over microwave - and that's with 18 repeaters.

I certainly am. I got my DK1 on schedule at a great price, AND I'm getting to see VR succeed in the marketplace. And as a bonus, I'm watching Oculus and Palmer do quite nicely out of it.

I don't remember "stick it to the Big Guys" being a campaign goal on the Kick starter pitch.

Contradiction detected. You want developers to stop building in high levels of detail, but then say their audience won't settle for anything less?

This is exactly what the detail slider is for. You can't really fault developers for making their game look even more awesome on future hardware while still being playable today.

Just turn the detail down if you need smoother play (in VR or not), and have a closer look at your apparent need to max all the sliders. Don't force the developers to artificially limit detail for everyone.

Back to only two possibilities again, status quo or mud huts? The world really isn't that black & white.

The great majority of which can be substituted with alternate elements that have a similar effect. For example, the gold on edge connectors could be replaced with any of the corrosion-resistant noble metals - silver, iridium, platinum, rhodium, titanium etc.

The infographic ignores undeveloped and undiscovered reserves, as I've said, so is no real guide at all. Extrapolating its claims to include data it does not show is pure speculation.

The linked study was informative, thanks. Interesting to see that many substitutions are indeed possible while for some, no practical alternative has been found yet. But bear in mind, for many of these critical materials, we simply haven't looked for an alternative yet, and some will likely be found when supply gets expensive enough to justify it.

For the remaining materials, as the study itself says, we can instead develop "new and transformative technologies, many of which are under active investigation: advanced composite materials, bulk metallic glasses, and structural biological materials, to name a few."

It's not unreasonable to expect that, given all the above alternatives and coupled with the future sources of improved recycling and asteroid mining, materials supply will likely be little more than speedbumps along the road of progress - as it has been through cycles of supply and demand for all history. I see little reason to give in to pessimism at this early stage.

I'd take any claims China makes about its rare earth quantities with a few tonnes of NaCl. They've been restricting their output to boost prices, and using "limited reserves" as an excuse - but they're certainly not the only source. There's plenty more reserves which are being opened up now that China's prices aren't so cheap.

The Visual Capitalist infographic is pretty, but is apparently based solely on current mines & sources, as far as I can tell. It mentions the existence of undeveloped and undiscovered reserves, but doesn't try to estimate depletion rates of those. While of course I wouldn't claim we'll "never run out", we can clearly go a lot further than the infographic shows before the price per unit extracted gets excessive - in most cases long enough to find alternate sources as I mentioned above (we've already started eyeing the asteroid belt).

Plus of course, few individual minerals are absolutely essential anyway. Most have alternatives that can be substituted, and demand for more than one mineral has waxed and waned as technology developed a use for it, then replaced it with something more effective. I looked up the USGS report on Antimony, for example, and it makes interesting reading.

Unsustainable over what term, though? There's plenty of most minerals around. Rare earths aren't particularly rare, and there's a lot more sources of most things if we spend a little more to develop them. And recycling pushes "peak minerals" out further. We're set for most things for the next few decades at least, centuries mostly, and millennia for a lot of common stuff.

In the longer term, we can greatly improve our recycling (nanotech molecular disassemblers combing our landfill, maybe), and our sources (there's literally astronomical amounts of useful minerals in the asteroids). I don't think we're in too much trouble there.

Read up on opportunity costs. Money isn't lost completely, but it could certainly be spent more productively.

Exactly. It's definitely true that e.g. renewable energy development would have many other knock-on benefits, but it's better not to have to adapt to dramatic climate change while we're doing it, any more than needing a World War to prompt us to invent radar.

Yep, the Thwaite Glacier vulcanism is definitely a factor in this - it's certainly not all due to climate change.

The accelerating sea level rise from this is still a problem regardless of cause. From your link:

Are those really the only two possibilities that occur to you?

No scientist claims the world is ending; that's a straw man from the denialist camp. What the scientists ARE telling us is that the coming climate changes (which can't now be prevented completely but CAN certainly be mitigated) will have significant costs - economic and humanitarian.

Even if you ignore the human costs (relocations, famine, refugees, conflict over dwindling local resources - mostly in poor countries), there's still the economic costs (increased storm damage, droughts, flooding, sea level rise) which been shown by numerous economic studies to far outweigh the costs of mitigation.

Yes, it will cost money to move our energy infrastructure away from fossil fuels. No, it won't derail the economy (the average estimate from the World Bank and many others is about 0.5-1% of GDP). But it will slow the onset of climate change, reduce the impact of changes in decades to come, and SAVE us hundreds of billions we'd otherwise need to spend adapting to the negative effects of dramatic climate change, not to mention other indirect benefits (like the surprisingly large health costs from fossil fuel pollution). Many studies show the investment in a clean, efficient energy infrastructure will actually save us money in its own right, independent of climate change effects.

Replacing lightbulbs is easy, low-hanging fruit, and there are numerous other efficiency gains we can make, but it's ultimately not enough. We don't have to cut our energy usage to the bone, we just have to invest in carbon-neutral energy generation - then we can easily support our lavish lifestyles with zero carbon cost, and save money in the process.

There are certainly other solar irradiance fluctuations greater than that, of course. GP was only referring to the long-scale growth in solar output.

The sunspot cycle varies output by about 0.1%, and of course there are Milankovitch cycles (orbital variations) which also affect the solar irradiance we get (though not the star's output). The Maunder minimum appears to have been due to an anti-phase correlation between the sunspot cycle and heliospheric current
sheet inclination variations (interesting paper here).

If you think the current climate research is "silly", have you considered that you may be simply misinterpreting it? Perhaps due to an incomplete picture, not having kept up with the vast amount of research in the field, or even by just taking out-of-context statements as meaning more than intended. And as for lack of supporting evidence, I don't think posts on slashdot should be taken as indicative of actual climate research. There's a lot of evidence out there if you look.

Because it's a feedback loop. Other factors (e.g. orbital variations) can initially trigger warming in the oceans, which then release CO2, which results in further warming. See this article for details.