Tech Tuesday – Superhero Origin Edition

Oscar Gordon

A Navy Turbine Tech who learned to spin wrenches on old cars, Oscar has since been trained as an Engineer & Software Developer & now writes tools for other engineers. When not in his shop or at work, he can be found spending time with his family, gardening, hiking, kayaking, gaming, or whatever strikes his fancy & fits in the budget.

Related Post Roulette

46 Responses

  1. James K says:

    Aero5: Torch drives, now we’re talking!

    Areo4: Its also possible that the big filter on technological species is behind us, not ahead. It may be that species intelligent enough to develop technology evolve very rarely – humanity got pushed through a very narrow bottleneck in our recent genetic history and it may be that it takes pressure that strong to lead to the combination of high enough intelligence and sociability required to produce serious technology.

    The trouble is that all discussions of the Drake equation and the Fermi paradox are workings with a single data point. There’s not a lot you can safely infer from n =1.Report

  2. Kolohe says:

    Aero1 – we already sent the ETs dick picsReport

  3. Michael Cain says:

    Military… You can change “I doubt you’ll see” to “You won’t see” for retrofits. The Navy’s already said that retrofitting EMALS to the Nimitz-class carriers would require a structural redesign, then tearing out and replacing significant chunks of the ships themselves. The analyses I’ve skimmed suggest that Navy brass already know it’s so expensive that they would have to give up one of their new Ford-class toys for each Nimitz-class retrofit.Report

  4. Kolohe says:

    As far as I know, the current Virginia class SSN block is still nuke to steam propulsion. There’s long been a desire to shift subs to all electric propulsion. There’s been one-off builds twice, the Tullibee and the Lipscomb, but nothing has stuck. Some reading this morning indicates there’s a push for all electric drive for both the Ohio replacement and the eventual SSN (X)

    Though even if these subs go to all electric drive, I’m fairly certain it will still be a primary cycle/secondary cycle pressurized water reactor that in turn uses steam to drive an electric turbine.Report

    • Oscar Gordon in reply to Kolohe says:

      Unless the Navy figures out a good way to extract power out of a nuke plant without making steam, there will be a steam turbine in there somewhere. But there are benefits to having the shaft turned by electric motors.Report

  5. veronica d says:

    [Comp1] The sad part is, the fact that we disallow race as a training variable for most ML algorithms means we cannot control for racial disparity in deciding outcomes.

    For example, the ML algorithm that was used (I believe in Florida) to decide recidivism rates, and in turn eligibility for parole, could not factor in race. However, it could factor in variables such as, “Are any close family members also criminals,” or “Does the subject live in a high crime area.” Now, it is easy to see how such variables would correlate with recidivism. However, they also correlate with race.

    Here is the thing: it is possible that the correlation between “criminals in family” is different between white people and black people. I don’t know, but I can easily see how it might be a stronger indicator for a white person than a black person. This could happen because black families could plausibly have developed better social tools for managing such situations compared with white families.

    Who knows. The answer is in the data.

    In any event, there is no uniform American, nor is white culture somehow the default. We might aspire to some kind of melting pot — and why not! But if black people are systematically disadvantaged in housing, employment, criminal justice, etc., well dammit a melting pot is a fantasy that will burst on contact with reality.

    Ignoring inequity is not the same as achieving a balance. In practice, it is the opposite.Report

    • Kimmi in reply to veronica d says:

      Not sure how you get “most”. Not sure you know most.
      Race is stupid, anyway. We have fucking computers, you can get more finegrained than that.Report

    • Oscar Gordon in reply to veronica d says:

      Algorithms need to have practice data so we can spot the deficiencies early and adjust. It’s disturbing that governments are simply adopting an algorithm as provided and deploying it as is.Report

      • Kimmi in reply to Oscar Gordon says:

        Um. I don’t think they are.
        But as for immoral algorithms? the NSA Likes Those A Lot.Report

      • Pinky in reply to Oscar Gordon says:

        Every law is just an algorithm. They’re just less complicated – or maybe more complicated. Ditto every regulation. That’s part of being a society of “laws, not men”. And it is a risk if laws aren’t revisited regularly.Report

        • veronicad in reply to Pinky says:

          No, a law is not an algorithm. An algorithm is precisely defined in a way that a law is not. The terms that appear in an algorithm are pure computation. They are merely terms, squiggles on an endless tape churning through your Turing machine. (Although actually, I prefer that lambda computation model, but it’s all the same really.) The tape begins with some set of symbols. You feed it to the machine, turn the crank, and out comes a new set of symbols. Bam!

          Laws do not work that way. They are specified semi-formally in natural language — granted it is a particular flavor of language with its own traditions. It is surely more formal that the truly informal version of English we (for example) use on this forum. However, it remains natural language, which is both produced and consumed by humans in light of our full cognitive capacity — although again, a lawyer has special training to interpret that language. It remains natural language.

          It is not an algorithm. Judges interpret. They can handle new situations. That can judge. Turing machines don’t work that way.

          (It is reasonable to ask whether deep learning networks really count as “algorithms” any more. Certainly they use algorithms, most importantly back propagation), but is the final trained network an “algorithm”? After all, it’s just a set of weights.

          It’s a fun question, actually. I tend to say, no, it’s not an algorithm. It’s something else.Report

          • Michael Cain in reply to veronicad says:

            When I was a legislative staffer, I used to say that much as economists are often described as having physics-envy, law-makers had formal-specification-envy. They want to write formal specs, but English is ill-suited to the job*, and the courts keep changing the meaning of the words. Much like trying to write programs for a processor where the meaning of “add a,b,c” is subject to change without notice.

            * When I wrote the occasional formal spec for a living, I tended to write documents with facing pages. On the left page was a piece of the spec written in a formal language; on the right was an English-language description; at the beginning of the document was a statement that in the case of conflict, the formal-language version had precedent.Report

          • Oscar Gordon in reply to veronicad says:

            A Fuzzy Algorithm? H. Beam Piper would be so proud.Report

          • Kimmi in reply to veronicad says:

            self-modifying algorithms aren’t quite as determinate.
            Isn’t it strange that AIs have personalities?Report

    • Kimmi in reply to veronica d says:

      I’m not sure why you’d disallow race in the first place, frankly.
      Who wants a pornography detection bot that can’t find black people? (or fill in whatever color you want).Report

  6. Pinky says:

    Aero4 – If we are typical, then we are unique, because there is nothing else like us. That’s roughly what’s being argued in the piece.Report

    • Kimmi in reply to Pinky says:

      you should see the other ideas. (They run off the “There’s too many planets for there not to be evolved life” AND “intelligent life would have created signs… unless”).

      Of course, the fun idea is Mirrorland, where most of the universe is just a vaguely sketched simulation. (so called because sending things outside the border of “actually modelled” makes them jump to the other side)Report

    • Oscar Gordon in reply to Pinky says:

      This is why I included it. It seems a bit too neat, and circular. Figured folks would have fun with it.Report

      • Kimmi in reply to Oscar Gordon says:

        It’s occams razor. Simplest, cleanest explanation.
        I really wouldn’t want to write the scientific paper trying to explain why and how Earth is under some sort of intergalactic interdiction.
        Or that every other intelligent form of life is being systematically snuffed out (or otherwise hiding for genetic reasons).

        We do have other explanations.Report

      • Pinky in reply to Oscar Gordon says:

        The principle of mediocrity is a reasonable idea, but it’s literally a conclusion based on a single datum. When you have only one observation, you shouldn’t disregard it, but you can’t generalize from it. “All coin tosses come up heads.”

        I just wrote a second paragraph, but I had to delete it, because I’m assuming that the article reasonably represents his thinking. I should know better than to make that assumption. The author wasn’t writing to scientists and mathematicians; he was trying to tell a story. He does so by having Whitmore tell two stories: one, about how we’d fit into a universe with a lot of intelligent species, and the second, about what happened to all the other intelligent species that should be out there. Neither is scientific. And neither do statistics nor the principle of mediocrity offer proof.Report

  7. North says:

    What I’d love to read is someone meditating past Aero4. What if we get a really solid idea that we are along in the galaxy? What’re the implications for humanity alone rattling around in an empty cosmos? Putting it in sci-fi terms what if we discover we are the precursor/ancient/elder species? What do we do with all that space?

    Or more pessimistically, what if all the quantum stuff stays quantum and we can’t break the light speed barrier? What if we’re not only alone but imprisoned mostly within our terrestrial gravity well but more absolutely within our solar system? Ugh.Report

    • Kimmi in reply to North says:

      if we’re alone in the galaxy, it’s a really, really small galaxy.
      And there’s not a lot of space.Report

    • gregiank in reply to North says:

      Barring new physics this solar system is what we got. Given we can’t even do that much here yet i don’t’ think we should worry about going to the stars. Let’s just see how many decades it will take to have a minimally dependent workable colony on Mars. That is a big enough task. We’re barely off of the Earth really.

      If humans are the precursor species then that makes our present time pretty much the stone ages of the Elder’s. We’re not even going to a memory to the Tau Cetians or Alpha Centruions.Report

      • Kimmi in reply to gregiank says:

        You mean we already created *muffles voice* and that doesn’t count?
        Sci-fi, dude, is full of stories… I can’t believe we haven’t created SOME new physics this century.
        (By which it should be clearly understood that I mean new engineering, as that’s what it takes to find … things that I’m not supposed to be talking about)

        We got the hard part down, how hard is warping spacetime?Report

      • North in reply to gregiank says:

        Yeah exactly, that’s along the lines of what I’m thinking too.Report

      • Oscar Gordon in reply to gregiank says:

        Alternatively, we are but one of a number of races coming of age at about the same time (i.e. it takes about 13B years for the universe to settle down enough for intelligent life to form and climb out of their own gravity well).

        Or, as one SETI scientist put it, we haven’t figured out how the space faring races communicate across the stars. Maybe they use Zeta waves, and once we discover Zeta waves, we’ll suddenly find evidence of aliens.Report

        • greginak in reply to Oscar Gordon says:

          Given the time scales involved i think it’s unlikely there would be multiple civilizations at similar levels of development. There may be other civ’s but they would be far more evolved or far behind even if there is a way to communicate or travel across the stars. And who knows if there is just one tech civilization per galaxy then we’re never gonna know.Report

          • Kimmi in reply to greginak says:

            Yeah. But given the number of planets per galaxy, it would be REALLY weird not to have more than one.
            And with self-propagating explorer bots, you can even cross from galaxy to galaxy. They aren’t THAT hard to make.Report

        • James K in reply to Oscar Gordon says:

          Maybe they’re all just pretending they’re not there because we’re made out of meat.Report

  8. Morat20 says:

    So, apparently scientists have come up with a better way to recharge zinc-air batteries.

    Which is pretty cool. Except for the fact that zinc-air batteries sound like magic. For the air. I can totally believe “zinc” is involved in a battery. It even sounds fairly electric and exotic, the sort of thing you’d use in anything nifty and scientific.

    It’s the “air” party that’s just flabbergasting. My lizard brain assures me that air is “nothing”, so “zinc-air” sounds like something from nothing.Report

  9. Kolohe says:

    USS Fitzgerald JAGMAN (or something like it) is outReport

    • Kolohe in reply to Kolohe says:

      Having now read it, it’s not about the investigation about how the collision happened, but about everything that happened after the collision. Some harrowing stuff when they describe the flooding and egress from the flooded spaces.Report

      • Oscar Gordon in reply to Kolohe says:

        You were Navy, right?

        Nothing gives you nightmares more than fire & floodingReport

        • Kolohe in reply to Oscar Gordon says:

          I was on subs. A breach like that would have been catastophic*. But maybe, LA and VA subs would have punched a hole in the cargo ship instead of vice versa.**

          *the closest thing the sub force got to something like this in recent memory was when the San Francisco hit a seamount at 20+ kts. The rapidly deceleration caused massive injuries (and one fatality). Hull integrity was maintained because the sonar sphere access hatch was properly dogged shut. (And the people asleep in their racks were mostly unhurt in this case, as most berthing is oriented along the fore/aft axis)

          **e.g. the Greeneville v Ehime MaruReport

          • George Turner in reply to Kolohe says:

            How come Navy ships have never used revolving doors? Large buildings use those because they operate just fine in the presence of significant pressure differences, as the pressure pushes on both sides of the axle evenly, and no open flow path is ever created despite people coming and going.

            They way you’d escape from a flooded compartment is just walk through the revolving door to a non-flooded one, just like you do every day, except with a small slug of water coming along for the ride.

            You could add a brake to the axle to lock any and all doors whenever you wanted, with perhaps a manual release. There would never be a state in which the ship’s watertight doors weren’t watertight because they’d always be watertight, even when in use.

            The space and weight requirements would be larger, but that might be an affordable trade off. Getting through them would be a bit slower, but that might be acceptable too, depending on how many there were in a given path.Report

            • DensityDuck in reply to George Turner says:

              Because if the deck warps under a swinging hatch, you can still open the hatch. If the deck warps under a revolving door you’ll never open it again.Report

              • George Turner in reply to DensityDuck says:

                Hrm… That’s a good reason, but one that a stout frame would likely prevent, though that’s more weight. I would also of course make most of the revolving doors elliptical, so they’re more like a half-hatch on each side. Sealing might be a bigger issue, especially considering wear.

                And of course the revolving door would itself end up weighing a lot, perhaps too much to easily push, in which case you’re adding a control system which will of course work fine unless the compartment is flooding.

                Or you could combine a lighter revolving door with a conventional hatch at one end, which normally stays open, but which is the final, high pressure seal in an emergency.

                One other advantage of revolving doors is that they’d also only pass any smoke along when someone went through them, which would be a good thing.

                Of course all this assumes that sailors can make it successfully through revolving doors. That might require some experimentation.

                And also on the down side is that no cords, wires, ropes, or fire hoses can be pulled through one. That might be a show stopper.

                Maybe they’d make sense on space stations, at least until we perfect the turbolift.Report

              • Oscar Gordon in reply to George Turner says:

                Revolving doors add too much needless complexity. Sailors understand that, while every sailor is valued, none are as valuable as the ship, so if the hatch is dogged and the water is rising, if you can’t reach an escape scuttle, you will die in that compartment, because the water can not be allowed to flood other parts of the ship.Report

        • Kolohe in reply to notme says:

          Commander Benson and the Fitzgerald’s second-in-command executive officer, Comdr. Sean Babbitt, both of whom were asleep in their cabins when the ships collided,

          This right here would be against standard procedure on any boat I’ve been on while operating they were. The shipping density there is high enough so if we had to night steam, awaiting a morning port entry window, the Executive officer was awake and stationed as Command Duty Officer.

          And even at that, there were distance to nearest contact tripwires in the standing orders that required the Officer of the Deck to wake up the captain anyway.

          Just based on this one factoid, looks to be shaping up to a case of being to comfortable (i.e.complacent) because you were only doing ‘local ops’.

          The report also indicates it was a clear night with reasonably calm seas, and the first inidication of trouble was the actual collision, so more precursors to complacency. Everything I’ve read indicates the cargo ship was squaking AIS, so even with a crazy ivan or whatever from the cargo ship, the two vessels should have never gotten that close.Report

          • Oscar Gordon in reply to Kolohe says:

            If the Crystal was squawking AIS, the CIC should have had them on the plot, and yeah, course changes should have been noted and logged (and reported to the bridge).

            Who did they have as OOD, some fresh faced Ensign or JG? And what was the Chief doing, playing Candy Crush on his phone?Report