Check out the floor of the Pantheon, in Rome. THAT IS THE ORIGINAL FLOOR. People have been walking on it for 2,000 years and it's still gorgeous.
The sheer competency of these guys was unbelievable.
10,942 karma · joined November 30, 2011
The sheer competency of these guys was unbelievable.
Given those kind of inputs, how can one say with any confidence when Oregon will peak, and how large that peak will be?
Other states have much better testing and thus can be more confident in their predictions. Some have worse. For most, however, I get the impression that this kind of modelling is somewhere between a wild-assed guess and a crapshoot.
Now, let's say I own an apartment block and want to advertise for tenants, but I don't want any of them dirty muslims in my building. I can target my advertising to exclude people whose interests include Halal food.
Or, let's say I'm the propaganda department of an adversarial nation state, and I want to stir up communal tension. All I have to do is find people with an interest in Halal food, and target them with ads exhorting them to join other Christians in the next Crusade against the Moors. Or whatever. This was actually a common technique in the 2016 election: identify voters who, via their social connections and profile interests, are likely on the fence about Clinton vs. Trump, and also are a bit homophobic -- then target them with "Queers for Hillary" advertisments.
It's shocking how easy it is to weaponize this stuff.
The real reason, per his telling, was actually very close to the official reason: the welds on the coolant intake failed. The coverup, however, was this: the welds had actually been made using a new, then-highly-classified technique (some variation of electron-beam welding, I think), which the grunts doing weld inspection didn't have clearance to know about. So they were told to skip the inspection on those particular welds. Which yielded predictable results
I've spent the last 25 years wondering whether I'd accidentally stumbled across something classified, or was just being bullshitted by a drunk at a party. I look forward to finding out!
1: https://www.canadianbusiness.com/the-last-days-of-target-can...
Nevertheless, here goes:
What won't change (much):
- We still won't have L5 autonomous cars.
- We will have L4 in various places, mainly as public transport systems, and this will essentially evolve out of existing segregated autonomous on-demand systms, such as the WVU PRT and the Heathrow Pod. Increasing sophistication of the navigation AIs will make the degree of segregation less and less over the years, but full automation will still require a significant element of environmental / infrastructure design, which is what makes it L4. There will be a lot more such systems, but it will be an evolutionary rather than a revolutionary change.
- No AGI, sorry. Not that we'd necessarily recognise it if we had it. There will still be no effective working definition of "consciousness".
- Laptops will still suck. Different constituencies will still regard old Macbooks or Thinkpads as the zenith of laptop desgin, with a nostalgia that resembles that of Amiga or BeOS fanatics. (And this won't make them wrong.)
- Linux will still be niche on the desktop. Although enthusiasts will claim that it doesn't, it'll still suffer from the same useability problems that have it did in 2020, and 2010, and 2000. "Oh there's some stupid mouse behaviour which is making your GUI almost unusable? You can change it with this simple hack of X-windows in the shell."
- Social media will still suck, and will still be a vector for significant psyops, both political and commercial, with politically and economically destabilising effects. Dictatorships will be better at weilding this than democracies, which will still be on the backfoot.
- The climate crisis will still be going from bad to worse, unfolding as a series of localised extreme weather events and population displacements. Although the sociopolitical response to it will finally be turning the other direction.
What will change:
- The majority of vehicles sold will be full-electric.
- The way people live will be changing. A lot more co-ops, co-living, and even straight-up communes. This will represent a very small percentage of the population, but the social effects will be significant.
- Universal Basic Incomes will have been implemented in several places.
- There will be dozens of people living on Mars, hundreds on the moon, and thousands in Low Earth Orbit.
Definitely recommended!
To amplify what this presentation says, here's the hard truths about predicting social outcomes: either they
1. Are simple and obvious, and can be easily understood and predicted via regressions and trendlines.
2. Are complex, non-obvious, and neither can nor should be predicted. In fact, attempting to predict them is often dangerously wrong. But this doesn't mean that they can't be understood.
In the first instance, you don't need any fancy software, so there's no snake oil to be sold. The second instance, however, gets caught in a cognitive bias: we think that the future can be predicted. This is because, in simple systems, it can: drop a glass above a hard floor, and you can accurately predict that it will fall and shatter. That is a simple system. We think that complex systems must be similar... just more complex.
But complex systems are fundamentally different. Consider a double-pendulum. Its movement can't be predicted for more than the next few swings. Even if you know the exact starting configuration of the pendulum -- and by exact, I mean not just every single sub-atomic particle in the pendulum arms, but the gravitational influence of literally every object in the universe -- you still wouldn't be able to greatly extend your foreknowledge of its movements. This is because the feedback loops are driven by chaos, and chaos is baked directly into the mathematical fabric of the universe itself.
For the mathematically inclined, consider the Mandelbrot set: it is, essentially, the equivalent of a double pendulum. It asks a question: "by starting with this number and iteratively exponentiating it, will in trend towards zero or infinity?". When you ask this question on a simple number line, then the answer is obvious: below 1, it trends towards zero; above 1, it trends towards infinity. However when you ask this question on the complex number plane -- with two numbers feeding back to each other as they exponentiate -- then the only way to answer the question is to keep iterating the numbers to find out. There's no shortcut. In some places, the answer is found quite quickly. In others, it takes thousands of iterations. In still others, it takes infinite iterations: you could build a computer the size of the galaxy and you still wouldn't be able to answer the question of whether such-and-such coordinates trends towards zero or infinity. That's the complexity you get from just two number lines and a very simple feedback loop.
The real world of psychological and social cause-and-effect is far more complex than a double pendulum or a Mandelbrot set, and attempting to predict it is even more futile. In fact, it's dangerous.
What makes this dangerous is that you can throw statistics at complex futures, and make statements like "Future A has a 40% probability; futures B, C, and D all have a 20% probability". But the human mind is terrible at making good judgements based on this kind of information. Hearing it, people tend to think: "right, that's settled then: Future A is twice as likely as any other scenario, so that's what we'll plan for." We fixate on what looks like the most likely future, and disregard the rest.
The problem is that if the preparations for Future A are contrary to what you'd want to do in Futures B, C, and D, then betting everything on Future A means that 60 % of the time, you'll lose.
What's interesting is that increasing the accuracy of those percentages doesn't necessarily help, and in many cases can hurt, since it only reinforces our tendency towards target fixation. In a 40/20/20/20 scenario, we might still make some concessions towards planning for the "20s". But in a 70/10/10/10 scenario, those "10s" will simply be discarded. Which means that 30% of the time you won't just lose: you'll be blindsided. Utterly fucked.
Unfortunately, most of the predictive AI that I've seen is focused on either increasing the "certainty" via extremely dubious means, or simply hiding the non-dominant answers altogether. So the AI does the target fixation for you. That's not a good thing.
The entire discipline of Scenario Planning[2] evolved to help people and companies "un-predict" the future. Rather than putting percentages on probabilities of outcomes, people need to understand the possibilities of outcomes. Even if complex systems can't be predicted, they can be better understood, and that can be very valuable for helping to navigate them in real-time. Rather than giving people the easy (and usually wrong) answer of "here's what is going to happen", you can give them a range of futures that could happen, and understanding those possibilities can aid navigation and lead to better outcomes.
It may even be that AI has a legitimate role to play in this process -- but it won't be in falsely predicting the outcome of complex systems, nor will it be in absolving people of the responsibility of thinking for themselves. This, unfortunately for my company's ability to raise capital, is a much less sexy sales pitch than the AI snake oil. (Although it does get us smarter and less annoying clients, so there's that!)
I had the pleasure of meeting Carmack a few times over the years at small aerospace conferences. He's both as true a geek and as much of a gentleman as you might imagine. I'm really looking forward to seeing what he does with AGI.
> Researchers have found a woman with a rare genetic mutation that has protected her from dementia even though her brain has developed major neurological features of the disease.
> She, like thousands of her relatives, going back generations, was born with a gene mutation that causes people to begin having memory and thinking problems in their 40s and deteriorate rapidly toward death around age 60.
Another example: scurvy. Known to be cured by lemons since 1601; took 200 years and the better part of two million lives for this news to be properly disseminated.[1]
This is part of the reason that it can take so long to comprehend a genuinely innovative new technology: it doesn't win the game that's being played -- it plays a different game entirely.
When airplanes came along, rail companies could quickly (and correctly!) judge that they would never be hauling coal and cattle and such. Therefore: airplanes are just a toy; we can safely ignore them.
Similarly, when television came along, the radio industry said: "people listen to the radio when they are working, driving, or drifting off to sleep. You can't do anything else when you're staring at a screen, so television isn't a threat."
What these industries failed to realise is that while the the new technologies were non-competitive on the use-cases that railways and radios had optimised for, the new entrants could win on metrics that the old industry had never thought to care about.
(All this is fairly standard Christensen disruptive-technology stuff, BTW.)
Anyhow, the thing about the legacy aerospace industry is this: they weren't playing the "let's build and fly rockets" game. They were playing the "let's maximise our funding" game. The rules of this game are as follows:
1. Spend all the money you get, otherwise your budget will be cut next year.
2. Spend even more money than you get, and do as little as you can with it, so that you can plead the necessity of a funding increase next year.
3. The more money you get, the more of a political base you can establish. The more of a political base you've established, the harder it will be to cancel the whole program.
Underlying all of this is the axiomatic assumption that the spaceflight market is essentially inelastic. Weather and positioning and communications and spy satellites need to be launched at more or less any cost, but any reduction in launch prices won't be matched by a corresponding increase in demand. Therefore: if you cut costs, you're just leaving money on the table and putting your political base at risk. Don't do it.
Typical aerospace management was playing this game exceptionally well. NASA's heavy-lift rocket, the Space Launch System, has spent $14B last time I looked, and is still years away from its first launch. Its launches will cost a billion dollars apiece. That's actually not easy to accomplish. If I gave you $14B and told you -- a person with only a passing interest in space stuff -- that I wanted a heavy-lift rocket in return, I'm quite you could do it. All you need to do is put out a tender for heavy-lift services, with payments made on a milestone/per-launch basis. Award these to several teams to keep things robust and competitive. Job done.
On the other hand, if I gave you $14B and told you to spend it on the appearance of serious rocket development work, I think you'd find it quite challenging. Spending that money without flying a damned thing takes exceptional skill. Hats off to any program manager who can accomplish this!
But when SpaceX comes along and decides that the rules they want to play by are "build and operate the most amount of rocket for the least amount of money", the rest of the industry simply couldn't figure out WTF game SpaceX was playing at. Most of the industry still hasn't figured it out.
One revolution which I follow quite closely is in spaceflight. Since the beginning, it had been done one way and one way only, and its way of doing things had become a goal in itself. Its exorbitant cost was lauded for its ability to create jobs and -- for more ambitious initiatives -- to foster international collaboration, since surely it was too expensive for any one nation to undertake on its own. Its resulting rarity and exclusivity created a kind of mystique around rocket scientists and astronauts. This mystique and these justifications became fundamental to the narrative of what spaceflight was about.
Then SpaceX came along and screwed it all up, and practically nobody noticed. "Bad example," you might say: "Elon Musk is fetishized by millions, and is hardly out of the news cycle for more than a few minutes". Which is true, but Musk's status as a celebrity is entirely orthogonal to actually comprehending what he's doing, and adjusting one's model of the world accordingly.
What SpaceX has done/is doing, is this:
1. Changed traditional aerospace engineering, management, and procurement practices, so that they were more focused on efficiently building/operating rockets than distributing work sites around a critical mass of congressional districts. Nothing at all heroic or sexy about this, but it reduced launch costs by a factor of 3 relative to the industry norm.
2. Developed 1-stage reusability, which is quite sexy, and reduced launch costs by a factor of 6.
3. Developed the Falcon Heavy, which reduced launch costs by a factor of 8 (for large enough payloads, of which there currently aren't any).
What I find remarkable about this is that although the sight of boosters landing themselves has become common place enough to no longer grab headlines, the industrial / political community is still very much at the "Do Not Think Aeroplanes Will Ever Compete With Railroads" stage of comprehension. The fact that it's made Musk into a celebrity is, frankly, totally meaningless and facile. What's slightly more meaningful is that this has enabled a startup to capture the majority of commercial launch services in the world, and make a profit doing it. But ultimately, what a 8x reduction in the cost of spaceflight should do is change the the entire nature of the spaceflight business: enable many new types of business cases and missions. This is a fait accompli, but it's taking a long time for the world to catch onto this fact.
If you browse through an archive of traditional industry magazines and forums and such, you'll find almost no mention of SpaceX's innovations apart from dismissal and derision. Typical themes are:
1. Private companies can't afford to develop launch vehicles. (Still being claimed for a couple of years after they first reached orbit).
2. SpaceX is losing money. (They're not.)
3. SpaceX is only making money because they're subsidized by the government. (They're not.)
4. Recovering a booster is impossible.
5. Reusing a booster is impossible. (This argument was still being made long after SpaceX had started flying reused booster.)
6. Rocket reusability will never be economically meaningful because [insert painfully tortured and ridiculous argument here]. (Said after SpaceX had captured the bulk of the world's launch services).
7. Private companies can't develop heavy-lift launch vehicles. (Said as recently as a congressional hearing last week, after the Falcon Heavy has been successfully flying for almost two years)
8. Look, astronauts should be the very best of the best of the best. We shouldn't just allow anyone into space. (Said ever since the very first space tourist booked a flight to the ISS in 2001).
The myopia here is really striking to me. But one thing I've noticed is that -- contrary to the article -- it mostly doesn't really come from a too-conservative Joe Public. Rather, it comes from well-regarded experts with a vested interest in their world not being upended. When SpaceX first successfully recovered a booster in late 2015, I found that my friends with no particular interest in space were able to comprehend the significance of that milestone far more accurately than executives from the old guard space contractors. And I genuinely think that the old guard didn't comprehend. They weren't publicly ignoring/rubbishing SpaceX, while privately doing intensive R&D in an effort to catch up. Privately, they were doing nothing differently. So they really didn't understand what they were seeing.
4 years later, they're vaguely beginning to understand what has happened, and starting to think about how they might catch up. The European Space Agency is funding paper studies of reusable rockets which look almost exactly like the Falcon 9. SpaceX's main US competitor, United Launch Alliance, is aiming to start doing trials of partial reusability in 5 years. China is ahead of both of them, already doing booster-recovery experiments very similar to the early Falcon 9 recovery trials in 2013.
But this a terribly muted reaction to what has already been a significant revolution in spaceflight. And anyhow, it is all going to be much too little, much too late -- because actually the revolution was just getting started, and matching the capabilities of the Falcon 9 is about to be the least of any competitor's worries.
SpaceX is now developing the Starship, which will be fully reusable and plausibly stands to reduce launch costs by a factor of 50-100 vs. the prior industry baselines. This will be a mind-bogglingly huge development: at those prices, asteroid mining is viable, and building Mars colonies is with reach of many NGOs, never mind governments. At those prices, the fact that we live in a solar system with almost unlimited resources starts to become a meaningful statement. Once the culture has absorbed the fact that this is really happening, I would expect to see thousands of people emigrating off-world every year, to pursue the many opportunities elsewhere. This will be a BIG DEAL.
And it's happening right now. Starship's engines have completed development and are now being optimised for mass production. The first test-bed flew earlier this year. The first full prototype vehicle is under construction and will make its first trans-sonic flight as soon as next month. They expect to hit orbit by next year. Musk is saying as early as next Spring -- and this is undoubtedly too optimistic -- but at the pace they've been going, late next year is absolutely plausible.
Moreover, "the pace they've been going" isn't a matter of speculative kremlinology: they've been practically open-source about the whole endeavour, with all the parts laid out in a field and assembled in full public view. As if to try to convince the world: "hey, wake up, this is actually happening."
But so far, awareness of this seems to be largely confined to cult-of-Musk fanboys, and their equal but opposite anti-fanboys. With very few exceptions, few in the industry have noticed that this is actually happening. Based on what I've seen with the Falcon series, I fully expect this state of denial to persist for at least 5-10 years after Starship reaches orbit.
My main issues with the XPS are:
- Only slightly quieter than a vacuum cleaner.
- Unless you want to use the fancy GPU in Linux, in which case it's louder than a vacuum cleaner and has a very short battery life. (I keep the GPU disabled)
- Camera is placed at the bottom of the bezel, which is an ideal location for observing your fingers on the keyboard, or if you give the screen a jaunty tilt, peering far up your nostrils. Serves no other purpose. I really can't fathom the process by which that camera got greenlit. I use an external clip-on camera instead.
A Mac would certainly be better. For the same amount of power and storage, it would also be 3x the price, and (in recent years) probaly less robust.
Oh yeah, that! Lost another five-ish hours to that, and eventually just decided to live with crappy scrolling.
Two decades later, and I still can't find as performant or useable a system as that.
What this experience has taught me is that computing in 2019 basically sucks. The problems with 2000-era Linux, as described in the article, are very similar to the problems with 2019-era Linux. External monitors are a particular pain point for me. I've got an HDPI laptop and I want to plug into an old non-HDPI era monitor. Doesn't work. Spend the next 10 hours poking around forums, trying weird XWindows options, installing Wayland, etc. Still doesn't work. Eventually, give up.
Windows 10 works marginally better. Both remain vastly inferior to MacOs.
I'm not saying that the grass isn't greener on the other side. Macs are regressing, but the grass isn't greener on either side. Let's stop pretending otherwise.
Also, that sauce depositing methodology is just not right. This robot does it much better: https://www.youtube.com/watch?time_continue=11&v=cc-ClpymK_Q