Statistically that isn't true; engineering PhDs have virtually zero unemployment rates and high salaries in industries.
Granted, not high enough salaries to justify the time spent: you don't get a good monetary ROI on the PhD. But you can easily land a job at all sorts of places, ranging from Google to quant shops to Lockheed.
Pulling off the "if employers like this emphasis" part would be interesting in itself. Attempting to found a new vocationally-oriented, for-profit university specializing in technology is not a new idea. That's the ITT model, and several of their competitors. But these degrees have never gotten much traction among employers. They aren't worthless, per se, but they aren't anywhere near the value of a regular CS degree from a respected university.
This is the kind of situation where I think the publish/no-publish type binary is suboptimal. If they did legitimately conduct this experiment (not faked data, etc.), but it was just a poor experiment, I'd rather have the results recorded somewhere, and then some meta-data of which studies are actually decent ones whose results you should give significant weight to.
This one might fall pretty far towards the "unreliable crap experiment" side of the scale, so maybe not a lot is lost by just axing it. But many things fall in a gray area in between. Just axing the few worst experiments doesn't really solve the underlying problem that we should distinguish two things: 1) whether a paper faithfully reports the results of an experiment that was actually conducted as described, with the data competently collected; and 2) whether this experiment actually succeeds in demonstrating any interesting results with statistical validity.
I do think #2 is very important, but my general bias in favor of transparency leads me to prefer publishing the experiment/protocol/data in any case, as long as they aren't outright faked. Maybe they shouldn't be published in a proper journal as a "result", though. I think what I'd like to see might require a bigger reconfiguration of how the results of scientific experiments are recorded and analyzed.
Imo, withdrawing papers makes sense mainly if there is indeed, "evidence of fraud or intentional misrepresentation of the data". Faked data doesn't help advance science, and should be purged from the record.
But merely questionable conclusions are another story. Science is a back-and-forth process: someone publishes a study purporting to show X, and then someone else criticizes their conclusions, re-analyzes their data, attempts to replicate it, etc. Then they publish their own conclusions, purporting to show not-X. Withdrawing the original study in this case doesn't make sense to me, if it was not fraudulent: we don't typically retroactively go into old journals and blank out the articles that have subsequently turned out to be wrong. We just write new articles with better analysis.
How did this paper book's publishers get a copy of the story? They seem to have put out an unauthorized limited-edition run of 25 copies in 1999, but from what source?
I could see it being a problem if he were alive: if someone doesn't want something released, and it gets leaked anyway, it can cause various kinds of negative effects for the person (unwanted attention, etc.). But he's been dead for years, so I'm having trouble seeing the harm here.
That's an interesting social difference I noticed after moving to Denmark. It's super hush-hush to get any kind of treatment for mental illness in the U.S., and many people avoid doing it at all because of the stigma. But here, someone will just casually mention in conversation that they were out of work for 3 months last year because they enrolled in a treatment program to treat their depression + drinking problem. The state paid for medical leave and provided a treatment program, it worked, and they went back to work 3 months later, and they have no problem disclosing that. It's just seen as a thing that can happen and should be properly treated, but otherwise no need to be ashamed of it.
There is some existing work on the same basic idea of massively parallelizing regex matching by doing all the NFA branches in parallel, but using a GPU. Now a GPU is not necessarily perfectly suited to this problem, but it does have the advantage of being a mass-market consumer product, which produces economies of mass-market scale that let the average GPU have a ton more processing units and RAM than this Automata processor does. Would be interesting to see if an NFA-specialized processor gets enough of a speedup to overcome the manufacturing advantage of just throwing a beefy GPU at the problem.
I guess if you're exchanging $100k or something it could get tedious, requiring movie-style briefcases full of bills. But is this a practical problem for most people? If you want to transfer $1000, that's just 10 bills, which fit easily in a wallet or envelope without even being conspicuous. I don't do it regularly, but I've exchanged cash in those amounts without it being much of a hassle, usually for mundane reasons (one time I was being reimbursed for a trip to Taiwan, and the organization just handed me $1000 in USD in person, because it was easier for them than trying to figure out how to do an international wire transfer).
Even more than not designed to be anonymous, it's specifically designed to have a global, completely public transaction ledger. That is more or less the core of the design. How do you have "accounts" without a central server keeping track of them? The Bitcoin solution is a public ledger that all clients agree on. This public ledger then has an update mechanism designed in a way that's intended to make it so you can add transactions to the public register iff you have the private key of the account the transaction is "from".
Of course, you can try to keep your RL identity from being associated with a particular account number, but Bitcoin's design makes no specific effort to help you do so: the transaction graph is public, for anyone to do any kind of analysis they want.
Well, what they're doing historically is running a vertically integrated silicon foundry / chip-design business, not simply doing those two things independently. So they usually make sure neither side does anything to jeopardize the other, and if anything actively works to maximize the success of the other half. That can, in some contexts, be more profitable than running them as independent businesses that treat third-party customers neutrally. It's also usually legal unless you have a monopoly. If Intel had a foundry monopoly, they could be found to be misusing it if they refused to do business with companies that competed with one of their other business areas.
The problem is that, as currently priced, there aren't any "upkeep fees for the utility, and some small profit bonus for the utility" subtracted. The feed-in credit is at retail pricing, even though it would make more sense to credit it at wholesale, like any other generation source.
You can see that most clearly with someone who ends up at a net 0 kWh usage. Even though they send power both directions over the lines, since line maintenance is paid for by a portion of the per-kWh fee, and they use net-zero, they don't pay any line maintenance.
One possibility would be to break out line maintenance into a separate fee, and charge it on gross bidirectional, rather than net volume. But then you'd need the meters to work differently. Just charging a flat monthly fee for feed-in customers is a way of approximating it for typical-sized users.
But the court's decision doesn't argue that. It argues that intention is irrelevant, and there is no privacy expectation in this case even if the files were accidentally or otherwise unintentionally made available.
1930 is also about the time many of those Norwegian civilian guns were manufactured. It seems every Norwegian I know has a few old hunting rifles they inherited from their grandfather, along with a primitive mountain hut to store them in.
The paper this press release is about doesn't seem to be online, but two papers from the past few months analyzing this GRAIL data (with some of the same authors) are available:
Stalin was pretty good at it! Pretty bad at a lot of other things, but crash industrialization of Russia at all costs, that he was pretty solid at. To the extent that rural areas of Russia nowadays have a lot of USSR nostalgia, because the glorious Soviet infrastructure is slowly crumbling and not really being maintained anymore.
You seem to have somewhere in your explanation skipped hydro, which is actually the main baseline source of power. Sweden and Norway provide the main "backbone" of the Nordic power grid, and here are roughly their sources of energy:
Slashdot Mirror
Statistically that isn't true; engineering PhDs have virtually zero unemployment rates and high salaries in industries.
Granted, not high enough salaries to justify the time spent: you don't get a good monetary ROI on the PhD. But you can easily land a job at all sorts of places, ranging from Google to quant shops to Lockheed.
Pulling off the "if employers like this emphasis" part would be interesting in itself. Attempting to found a new vocationally-oriented, for-profit university specializing in technology is not a new idea. That's the ITT model, and several of their competitors. But these degrees have never gotten much traction among employers. They aren't worthless, per se, but they aren't anywhere near the value of a regular CS degree from a respected university.
If you find more wonder in the Discovery Channel than a good theoretical CS or physics class, you might have a superficial idea of wonder. :)
This is the kind of situation where I think the publish/no-publish type binary is suboptimal. If they did legitimately conduct this experiment (not faked data, etc.), but it was just a poor experiment, I'd rather have the results recorded somewhere, and then some meta-data of which studies are actually decent ones whose results you should give significant weight to.
This one might fall pretty far towards the "unreliable crap experiment" side of the scale, so maybe not a lot is lost by just axing it. But many things fall in a gray area in between. Just axing the few worst experiments doesn't really solve the underlying problem that we should distinguish two things: 1) whether a paper faithfully reports the results of an experiment that was actually conducted as described, with the data competently collected; and 2) whether this experiment actually succeeds in demonstrating any interesting results with statistical validity.
I do think #2 is very important, but my general bias in favor of transparency leads me to prefer publishing the experiment/protocol/data in any case, as long as they aren't outright faked. Maybe they shouldn't be published in a proper journal as a "result", though. I think what I'd like to see might require a bigger reconfiguration of how the results of scientific experiments are recorded and analyzed.
Imo, withdrawing papers makes sense mainly if there is indeed, "evidence of fraud or intentional misrepresentation of the data". Faked data doesn't help advance science, and should be purged from the record.
But merely questionable conclusions are another story. Science is a back-and-forth process: someone publishes a study purporting to show X, and then someone else criticizes their conclusions, re-analyzes their data, attempts to replicate it, etc. Then they publish their own conclusions, purporting to show not-X. Withdrawing the original study in this case doesn't make sense to me, if it was not fraudulent: we don't typically retroactively go into old journals and blank out the articles that have subsequently turned out to be wrong. We just write new articles with better analysis.
How did this paper book's publishers get a copy of the story? They seem to have put out an unauthorized limited-edition run of 25 copies in 1999, but from what source?
I could see it being a problem if he were alive: if someone doesn't want something released, and it gets leaked anyway, it can cause various kinds of negative effects for the person (unwanted attention, etc.). But he's been dead for years, so I'm having trouble seeing the harm here.
That's an interesting social difference I noticed after moving to Denmark. It's super hush-hush to get any kind of treatment for mental illness in the U.S., and many people avoid doing it at all because of the stigma. But here, someone will just casually mention in conversation that they were out of work for 3 months last year because they enrolled in a treatment program to treat their depression + drinking problem. The state paid for medical leave and provided a treatment program, it worked, and they went back to work 3 months later, and they have no problem disclosing that. It's just seen as a thing that can happen and should be properly treated, but otherwise no need to be ashamed of it.
There is some existing work on the same basic idea of massively parallelizing regex matching by doing all the NFA branches in parallel, but using a GPU. Now a GPU is not necessarily perfectly suited to this problem, but it does have the advantage of being a mass-market consumer product, which produces economies of mass-market scale that let the average GPU have a ton more processing units and RAM than this Automata processor does. Would be interesting to see if an NFA-specialized processor gets enough of a speedup to overcome the manufacturing advantage of just throwing a beefy GPU at the problem.
I guess if you're exchanging $100k or something it could get tedious, requiring movie-style briefcases full of bills. But is this a practical problem for most people? If you want to transfer $1000, that's just 10 bills, which fit easily in a wallet or envelope without even being conspicuous. I don't do it regularly, but I've exchanged cash in those amounts without it being much of a hassle, usually for mundane reasons (one time I was being reimbursed for a trip to Taiwan, and the organization just handed me $1000 in USD in person, because it was easier for them than trying to figure out how to do an international wire transfer).
That's true, but if you're trading physical tokens in person with people, you could just as well be exchanging USD.
Even more than not designed to be anonymous, it's specifically designed to have a global, completely public transaction ledger. That is more or less the core of the design. How do you have "accounts" without a central server keeping track of them? The Bitcoin solution is a public ledger that all clients agree on. This public ledger then has an update mechanism designed in a way that's intended to make it so you can add transactions to the public register iff you have the private key of the account the transaction is "from".
Of course, you can try to keep your RL identity from being associated with a particular account number, but Bitcoin's design makes no specific effort to help you do so: the transaction graph is public, for anyone to do any kind of analysis they want.
Well, what they're doing historically is running a vertically integrated silicon foundry / chip-design business, not simply doing those two things independently. So they usually make sure neither side does anything to jeopardize the other, and if anything actively works to maximize the success of the other half. That can, in some contexts, be more profitable than running them as independent businesses that treat third-party customers neutrally. It's also usually legal unless you have a monopoly. If Intel had a foundry monopoly, they could be found to be misusing it if they refused to do business with companies that competed with one of their other business areas.
Yeah, but I don't think I'm allowed to bring my brother to the exam?
That's only a portion of the maintenance fee: there is a fixed connection fee, and then a portion of the kWh tariff goes to maintenance as well.
The problem is that, as currently priced, there aren't any "upkeep fees for the utility, and some small profit bonus for the utility" subtracted. The feed-in credit is at retail pricing, even though it would make more sense to credit it at wholesale, like any other generation source.
You can see that most clearly with someone who ends up at a net 0 kWh usage. Even though they send power both directions over the lines, since line maintenance is paid for by a portion of the per-kWh fee, and they use net-zero, they don't pay any line maintenance.
One possibility would be to break out line maintenance into a separate fee, and charge it on gross bidirectional, rather than net volume. But then you'd need the meters to work differently. Just charging a flat monthly fee for feed-in customers is a way of approximating it for typical-sized users.
But the court's decision doesn't argue that. It argues that intention is irrelevant, and there is no privacy expectation in this case even if the files were accidentally or otherwise unintentionally made available.
Just as long as we get a year of the Linux desktop, I'm willing to settle for it at this point.
1930 is also about the time many of those Norwegian civilian guns were manufactured. It seems every Norwegian I know has a few old hunting rifles they inherited from their grandfather, along with a primitive mountain hut to store them in.
I have a can of Natural Ice here, and have a similar question to ask.
The paper this press release is about doesn't seem to be online, but two papers from the past few months analyzing this GRAIL data (with some of the same authors) are available:
"Gravity Field of the Moon from the Gravity Recovery and Interior Laboratory (GRAIL) Mission", the initial report of the observations
"The Crust of the Moon as Seen by GRAIL", reconstructing the crust thickness and composition from the observations
Are there really people who almost cannot function in the absence of a cell phone?
Stalin was pretty good at it! Pretty bad at a lot of other things, but crash industrialization of Russia at all costs, that he was pretty solid at. To the extent that rural areas of Russia nowadays have a lot of USSR nostalgia, because the glorious Soviet infrastructure is slowly crumbling and not really being maintained anymore.
hydrocarbons and nuclear
You seem to have somewhere in your explanation skipped hydro, which is actually the main baseline source of power. Sweden and Norway provide the main "backbone" of the Nordic power grid, and here are roughly their sources of energy:
190 TWh: hydro
55 TWh: nuclear
15 TWh: biomass/waste
10 TWh: fossil fuels
5 TWh: wind