This is so close to my sentiment that I had to double check to see if I wrote it.
And to explore running studies a bit further. The second time you build a system goes so much better because you already know all of the weird edge cases. And the third better still because your failures in the second time has cured you of some of your hubris.
Even if you somehow bankrolled 50 repeat projects and did the statistics etc correctly, you're still going to get some weird artifacts because some of those teams have people who did the thing before. You'll learn the wrong lesson when the real lesson is "make sure Bob is working on Bluetooth because he's done it 10 times before."
Starting with people with no experience is likewise not interesting because nobody really cares what lessons you learn by turning a bunch of muggles loose on something difficult.
What you need to bankroll is 50 teams worth of people who spend their entire careers testing out a hypothesis. (And even then you probably need to somehow control their professional communities in some way because again who cares what some small group of people approaches a problem when you could instead have people who go out and learn things from other people.)
Recently, I've kind of been wondering if this is going to turn out to be LLM codegen's Achilles heal.
Imagine some sort of code component of critical infrastructure that costs the company millions per hour when it goes down and it turns out the entire team is just a thin wrapper for an LLM. Infra goes down in a way the LLM can't fix and now what would have been a few late nights is several months to spin up a new team.
Sure you can hold the team accountable by firing them. However this is a threat to someone with actual technical know how because their reputation is damaged. They got fired doing such and such so can we trust them to do it here.
For the person who LLM faked it, they just need to find another domain where their reputation won't follow them to also fake their way through until the next catastrophe.
This is a fascinating idea, imagine a company spins up a super complex stack using llms that works, becomes vital. It breaks occasionally, they use a combination of llms, hope and prayer to keep the now vital system up and running. The system hits a limit, say data, code optimization, or number of users, and the llm isn’t able to solve the issue this time. They try to bring in a competent engineer or team of engineers but no one who could fix it is willing to take it on.
If they could do it digitally then you pull in your expert to testify that it was changed and they pull in their expert to testify that it wasn't or if it was then it was done by some mysterious hacker and they shouldn't be held liable for the attempt.
Meanwhile with paper, the judge and jury can look at your copy and see that it was obviously printed out and the aged for 10 years where their copy is fresh off the presses and your signature on it is either not a match at all or is suspiciously identical on every line.
Even if the digital version can be technically safer, I feel the paper version tells a more compelling story that non-cryptography experts can follow along with.
Some types of complexity have well formed mathematical definitions. Cyclomatic, time and space, kolmogorov, etc.
Something that's complex to the intellect is ill defined. I suspect it can be defined, but it's very highly dimensional. That is there is no scalar complexity number that can be used to compare items directly, and instead you may have two solutions of equal complexity but in different (and otherwise incomparable) axis.
The issue we have today is that we have no academic way to approach complexity so people just sort of develop an intuition for the subsets that they encounter and get bitten by.
However, because they have only a partial picture, they end up simplifying the complexity they do understand in a way that increases the complexity that they do not understand (at least until it's too late for the project).
I grew up on star trek TNG. However at a certain point in the past I was having kind of a hard time rewatching episodes. "We have the Internet and social media now, and they're obviously not going anywhere so why doesnt star trek have either? It is simply scifi of the past and now we need new scifi to incorporate new technological and social advancements."
These days though. Yeah, it's kind of obvious that you can't have a space faring civilization with the Internet and social media weighing you down. Honestly the Eugenics wars probably get kick started by social media.
A lot of Star Trek writing wildly errs with computers. And other things, but also computers.
Like, IRL we can't fire modern artillery over the horizon without a computer assisting us, and that's only a few hundred miles; a starship within range of their transporters (up to three times the diameter of this planet) is just an invisible dot on an invisible dot if you're looking for it out of a window. (IRL you can see the ISS flybys because it's only a few hundred km up, last I heard nobody can see any of the geostationary satellites).
Or comms: Uhura was written in an era when telephone switchboard were still around, manually connecting your phone calls by plugging and unplugging cords. (Did any later shows even have a comms officer?)
Even later, VOY tries to show how fancy the ship is with "bio-neural gel packs", but even when that show was written, silicon transistors were already faster (by response time) than biological synapses by the same degree to which going for a walk is faster than continental drift.
I meant more that the maximum range of an "over the horizon" artillery system is that.
I may have overestimated the maximum range even then, but the core point was that you need computer assistance even for relatively short distances on the ground, let alone in space.
How would the Internet work with interstellar distances? Even at Mars distances the latency to Earth makes it almost impossible for all but forums and email.
They can obviously communicate with Starfleet. "Subspace frequencies" or whatever they called it. Presumably personal and not just official communication would happen the same way. It's just not something that was top of mind when those shows were made. Long distance phone calls were still something you paid for at a substantial cost per minute. The idea that you'd be casually chatting with friends light-years away just didn't occur to anyone.
Not always. Lots of episodes they are well out of range of communications with Starfleet. They have even mentioned not getting a response from Starfleet for weeks in a number of episodes.
Except in Voyager (where due to the nature of the premise they had to be _somewhat_ consistent) this was entirely plot-driven. It could be anything from "real-time comms halfway across the galaxy" to "we'll get a reply next week" to "no contact at all". Occasionally this even varied within the same episode.
Which would be a welcome improvement. The speed of communication and content needs to slow down, and people need to return to longer form reading. People who lacked the patience and impulse control for this would actually drop off the platform, which would be a net improvement.
Presumably the same way faster than light travel works. I suppose you would wrap the IP packet in a warp bubble.
Or maybe the old adage of "a station wagon hurtling down the highway has more bandwidth than the biggest network links" would apply here -- send little storage modules at warp speed around the universe.
But also, in the show, they have clearly solved this problem, given that they can be out in Beta quadrant and still have live conversations with Starfleet back in San Francisco.
A side note, but there've been a couple of times in the most recent season of (the largely excellent) Strange New Worlds where I've thought "they're talking to the computer like an LLM now". The holodeck episode springs to mind, but I'm sure it's happened a few times.
I've noticed my mind thinking along similar lines when watching most recent movies. Many of the story points are driven by plots that would be upended if any one of the protagonists (or antagonists) had access to even the most basic of internet and/or portable communication devices.
Don’t they also have ways of sending messages wirelessly in real time, just bounded by speed of light? That’s a down-sight lot better than what we have now as we basically just blast radio signals in all directions at roughly the speed of light- which degrades very rapidly over distance.
I’m coloured largely by Voyager, but I don’t see any technology that we have now that they don’t have, not at the distances it would need to work at and without the infrastructure to make it work.
Star Trek has messages faster than the speed of light. And TNG and later have universal P2P communication with or without a reliable computer time-delaying it.
Honestly, I don't know what the conversation is about either.
Sorry, my musings were more general, not restricted to Star-Trek/... content. I mean more generally any movie in a setting within the last ~5-10 years~ to any time in the future. The fact that half the main characters / background extras / don't have their heads buried in mobile phones is by itself noticeable to me :)
Even in Star Trek weren’t the Eugenic Wars only ended by the invention of the warp drive (by a single guy no less) followed by first contact with an advanced and benevolent alien race?
No the Eugenics wars were endeded well before the nuclear holocaust of WWIII. When you see Zephram Cochran in First Contact, it was many years after Earth was devastated by nuclear winter. That's why populations were sparse, and there were various warring sects all over the world fighting for power. The warp drive (and discovery of aliens) is what united humanity after all the wars.
Oh, I get it. They're saying you should be able to write it in C.
[Jokes aside, I would be interested in hearing from bridge builders, aerospace engineers, or nuclear scientists. Pretty sure they're using math and not 'plain language'.]
And to explore running studies a bit further. The second time you build a system goes so much better because you already know all of the weird edge cases. And the third better still because your failures in the second time has cured you of some of your hubris.
Even if you somehow bankrolled 50 repeat projects and did the statistics etc correctly, you're still going to get some weird artifacts because some of those teams have people who did the thing before. You'll learn the wrong lesson when the real lesson is "make sure Bob is working on Bluetooth because he's done it 10 times before."
Starting with people with no experience is likewise not interesting because nobody really cares what lessons you learn by turning a bunch of muggles loose on something difficult.
What you need to bankroll is 50 teams worth of people who spend their entire careers testing out a hypothesis. (And even then you probably need to somehow control their professional communities in some way because again who cares what some small group of people approaches a problem when you could instead have people who go out and learn things from other people.)