Slashdot Mirror
Ask Slashdot: After We're Gone, the Last Electrical Device Still Working?
Leomania writes: After watching a post-apocalyptic Sci-Fi short on YouTube (there are quite a few) and then having our robot vacuum take off and start working the room, I just wondered what would be the last electric/electronic device still functioning if humans were suddenly gone. I don't mean sitting there with no power but would work if the power came back on; rather, something continuously powered, doing the task it was designed for. Are we talking a few years, decades, or far longer?
A very similar question was in an XKCD "What If?", but only in the printed book version (which has a bunch of extra chapters compared to the blog): "What would be the last artificial light source to glow when all humans were gone".
IIRC, the conclusion was that it would be status LEDs on space probes or radiation glow from buried nuclear waste.
andrew trigdell told me an amazing story back in 1999 about how he helped install Linux 0.99 on a solar-powered data collection computer in antarctica. Linux 0.99 was known to be highly stable, hence why it was chosen. it has a 56k modem which is enough to get the data back, and to check (very slowly) that it's still operational. so i think anything that's designed for long-term with those kinds of harsh remote and inaccessible conditions in mind, powered off of sustainable independent power, would be a good candidate for a device that would still be functioning even decades later.
Nope. It would scram when the rest of the electric grid collapsed a few days in. The plant has to constantly output power. When the grid fails, the plant automatically goes into safe mode to avoid tearing apart the turbines. Diesel generators then start up to run the plant until grid power returns but they'd only last as long as the fuel.
Nothing associated with the public electric grid would last long without humans present.
Moderating "-1, Disagree" is simple censorship. Have the guts to post your opinion.
The 10,000 year clock? If they ever build it.... Although, there's certainly a chance that a satellite sufficiently high enough in altitude with durable solar panels etc. would stay in orbit much longer than 10,000 years, if not functional due to it's batteries going dead. Perhaps there's already a satellite up there that will turn back on if it's panels are exposed to sunlight even if it's batteries are dead. I wouldn't know.
Hectice, baby, Mercator says hello to you
It is funny that people believe that we have, or will have "A.I" soon. There has been no progress towards TRUE A.I since the 1970's. ZERO. What you think is "A.I" is not really A.I.
Hint: Siri is not A.I, or a form of A.I, or even a precursor of A.I.
We don't design LEDs into our own boards, and we explicitly remove them from COTS boards that we use. Generally speaking the diffusers on LEDs outgas, meaning a) they are depositing materials on your spacecraft surfaces (bad) and b) could result in a shorting risk (also bad). There may be space-grade LEDs that big-space (think Hubble, JWST, Voyagers, etc.) use but I would be surprised. There's simply no need.
"Is it plugged in? Is it turned on? Is it on frequency?" solves about 99% of basic device connection issues. An LED will make one very short portion of that slightly shorter, and then only when testing on the bench, since you can't see it as soon as you box it up. As soon as you can talk to a device, you are able to run a long form functional test on it, exercising every part of the design and ensuring everything is working correctly. If it passes, you're good. If it fails, you pull the unit.
For ground support equipment, yeah sure, throw an LED on every rail and switch output.
Spoken like someone who knows nothing about AI outside sci-fi.
There has been a ton of progress towards 'true' AI over the last 40 years, it just has not managed to produce one. Progress != Success, esp on such difficult problems.
I don't know - a lot of current weak AI may end up being sub-systems for a strong AI - so in that sense we may well be getting closer. The problem is we have no real idea what strong AI might actually entail, implementation wise, and so have essentially no idea what if any progress we're making in that direction. At the very least we've found a great many strategies that don't work, which is in fact it's own kind of progress.
Honestly though I'm happy with the current state of affairs - weak AI may be able to get us into a lot of trouble (market crashes due to HFT algorithms anyone?) but it's nothing compared to what a strong AI would be capable of.
--- Most topics have many sides worth arguing, allow me to take one opposite you.
The former Soviet Union built hundreds of automated lighthouses in remote locations powered by radioisotope thermoelectric generators. Those use 90Sr, which has a half-life of 30 years so they can go for many decades. They were installed in the 1970s-90s, so most of them are around one half-life out. They could well continue operating for several decades, but some small solar-powered devices might well outlast them if they aren't damaged too badly by weather over the years.