Spoken like a man!
The potter’s wheel is commonly cited as the precursor of The Wheel (and still cited as the source for words like ‘wheel’ by the ever-diminishing minority that thinks PIE split up before wheeled wagons were in use). Wheeled toys for children were in use before wheeled carts, IIRC. As discussed in a recent thread, the wheeled wagon was not a trivial invention; whether you use a rotating axle or fixed axle you still need connections with just the right amount of friction. The pivoting axle (to aid steering) and the spoked wheel (for high-speed war chariots) were also key inventions made long after the wheeled wagon itself.
Beer was invented thousands of years after the earliest agriculture. However I have read the claim that the TRB-N culture (centered near Denmark and the last major “holdout” of hunter-gatherers) finally adopted a neolithic economy because they learned the recipe for beer! This may be a joke based on the penchant for beer-drinking by present-day Nordics, but there is archaeological evidence that the early TRB did drink mead and/or barley beer.
The TRB people are rather unique in that they were a mesolithic people who developed their own neolithic culture rather than adopting (or being overrun by) a neighboring neolithic culture. The independent status of the TRB culture may help explain the prevalence of I1 haplogroup in Western Europe areas dominated by the Bell Beaker haplogroup, as well as the unique (non-PIE) features of the Germanic languages.
Except Australia was colonized via South-east Asia, not sub-Saharian Africa.
This is, in my informed opinion, incorrect in two ways.
-
While progress has been accelerating overall, in some areas, we have been stuck against soft physical limits for a while. We know we can do a ton better, we just can’t do a ton better for a reasonable cost. Faster cars and aircraft and medicine are example of this.
-
Why are we stuck against soft limits? Because at the end of the day, our brains are at best only slightly improved from those of our distant ancestors, who barely kept it together just wandering around looking for food as nomads. Computers are only able to augment our brains in limited topics that are easy to program at the moment.
So, no. The ‘real’ singularity is the point at which we find a way to bypass #2. Whether we bypass it by developing a computer as smart as humans, or some cybernetic implant that makes humans 10 times smarter, or we find a way to copy a human to a computer, this will cause a hard edge. A noticeable jump.
No matter the method, since it took intelligence to develop the method to augment intelligence, the next step will come faster still, and so on and so forth.
Progress would accelerate until technology is approaching the real physical limits for technology. It would then decline slowly to zero, with tech now as advanced as physics will allow.
What are these real physical limits? We don’t know as humans now. But we can guess.
(1) For intelligence, a being can exist that thinks human-equivalent thoughts, but somewhere around 1 million to a billion times faster than you do.
(2) It is possible to build a set of machinery that can copy themselves almost as fast as bacteria do, but using rocks and solar and nuclear power.
So due to a set of doublings in (2) the end state would be all the rocks orbiting the sun in our solar system would have been used as feedstock for this machinery in order to make whatever it is that really high end intelligence wants to do.
Either way, the real singularity would have a beginning, middle, and end. We don’t know how long it would take but it probably would be concluded within less than a human lifetime. Though, due to (2), it is not known if any humans would be alive at the end of it to witness the results.
Yes and no, but to the degree that it’s relevant, mostly no. The brain of a modern human is basically no different from the brain of a Cro-Magnon, and very little different (and possibly even inferior to) the brain of a Neanderthal, Denisovan, or Floriens. But modern brains are far, far superior to the brains of our precursors. The difference is the use of the plural, there: Language and writing have enabled multiple brains to work in concert on the same problem, and so to come up with solutions that no single brain could find on its own. These are the two revolutions that have enabled us to bypass the limits of a brain. The computer revolution, great though it is, pales in comparison, and the biggest advantages even there have been from facilitating communications on ever-greater scales.
Chronos, agree or disagree :
- The fundamental circuitry the brain uses is still dog slow (a kilohertz at best, and synapse changes take much longer)
- Human brains are fundamentally flawed and contain many defects, making humans irrational in many common and uncommon ways.
- Human brains rapidly lose information over time
- It takes 20-30 years to train a human brain to peak performance
- Human brains lose performance with age
- Human brains fail and lose all data once a mere 70-120 years pass
Everybody agrees that greater worldwide communication, better spreading of knowledge, and language improvements all help. But they do not “bypass” any of these limits - they are still holding us back. They just allow us to have gotten this far despite the limits acting like a boat anchor.
Agree or disagree : if you could produce some kind of artificial brain that was not subject to #1-6, it would cause a rapid change in our world that dwarfs any other invention in all of history.
*Assume, for the sake of arguments, that the artificial brain runs 1 million times faster, is nearly perfectly rational, loses information but only when new information appears better than old, can load knowledge and skills from other artificial brains within seconds, can be copied when it’s hardware begins to slow and so always run at peak performance, and can be restored from a recent backup if the hardware ever fails completely.
Another way to think of it is that a human can only lift so much weight or run so far and takes x number of years to reach peak physical performance, which then degrades after y number of years.
Now imagine if we could somehow harness the power of larger animals like horses or oxen?
Or even better, imagine we could create some sort of artificial beasts that were orders of magnitude faster, stronger and more powerful?
Yes, human brains are individually slow, but they’re massively parallel, by a factor of over seven billion and counting. And brains only lose information if it’s not written down. Yes, if we could increase the performance of individual brains, that would increase our collective cognitive abilities, but it’s not the only way to do so.
This is the real sticking point. You can go back before WWII and find predictions of this, most of them assuming atomic power but treating it as magic. Unlimited, free power that came out of nowhere and had no downside. You can also find solar and wind power predictions from the 1920s that just … assumed.
No such magic power source will ever exist. If the singularity depends on one, then we won’t have a singularity. Or else it will always be 30 years in the future, like fusion power.
All other predictions depend on our learning what the hell consciousness is and how a brain works. Until then, any talk of limits on soft or brain power are meaningless.
I would be wary of dismissing solar power. Human civilisation uses less than a trillionth of the output of the Sun; we only have to sequester another trillionth of that power and we’ve doubled the energy available for our purposes. I think we can do much better than that.
But you are right in many other ways; the so-called Singularity is likely to be a sigmoid curve rather than an exponentiating one. But I expect we’ll go quite some way up the curve before it starts to level off.
Well, yes… “we hold these truths to be self-evident…”
There was a wave of settlement out of Africa, east along the south coast of Asia, and on to Australia. This happened about 60,000 years ago and assorted studies support it. Based on geography, we presume this migration involved people capable of building moderately good boats to traverse the various islands of Indonesia. This was followed by subsequent mixing and matching as waves went north/west to Europe, north and east into the Asian interior, back down to the rest of Asia to mix/displace earlier migrations, etc. But it appears these subsequent migrations never got to Australia.
Coastal travel by boat explains a lot about the speed by which humans could expand into new areas; plus boats had the practical effect of allowing fishing for food. Fishing was less likely to result in depletion of local resources the way hunting larger animals might. Yes, humans covering only 10 miles a year would expand pretty far in only 1,000 years - but desolate landscapes lacking in game or plants - deserts, mountain ranges - may be traversable by hunters, but present a far more formidable barrier to wandering tribes (breeding groups) than coastal voyages
Again, everyone uses the excuse “climate change” to explain why humans did not “invent” agriculture until the retreat of the last ice age 12,000 years ago. Yet, there were certainly areas of the world, perhaps further south, that were fertile enough and capable of supporting agriculture before that, during the ice age.
Every technical advance spreads when the advantage is obvious. Note how quickly horse use spread among the plains Indians once introduced… but not so much in less open spaces. Pretty much every human group knows about bows and arrows - the advantage in hunting is significantly greater. As discussed in an earlier thread - wheels are not useful without the pathways that make them useful. Carts for humans are useless unless ground is hard-packed; you need draft animals to overcome rolling resistance. Construction of wheels with the durability to carry heavy loads is not trivial and requires significant wood-working skills. So horses or oxen were a prerequisite to using the wheel in any significant amount.
Each level of additional civilization builds on the previous, from agriculture to settlements to the additional trappings of civilization. . Agriculture works if the plants and fertile areas exist. Animal herding is more flexible. The initial large worsk were stone - but stonework requires time and patience (and skill). Writing works if people have the leisure to learn, plus to read, plus places to keep a large store of these documents, plus the time and resources to make clay tablets or papyrus pages or whatever. Metal-working requires significant fuel plus a workshop, etc. As each skill becomes more specialized and precise, more resources are needed to support training the craftsmen to that level. All these extra tasks start with “need a food surplus”.
So the lack of anything else all boils down to the lack of agriculture - that’s the one main question about “why nothing until 10,000 years ago?”
True for the average peasant. But agriculture enabled a few lucky people to spend their entire lives doing something besides procuring food. Politicians, artists, engineers, architects, philosophers and all sorts of craftsmen owe their careers to agriculture, just as sure as the farmers do.
It doesn’t depend on that? I was trying to make safe, reasonable assumptions as to what a being that would be individually smarter than the entire human race is collectively could reasonably engineer, given the ability to experiment.
I don’t know precisely how the equipment would work, just that since we have proof of concept of all of it, I do know that at least one way would work.
Maybe it’s a tank sized machine that picks up moon rocks, scans them, and throws the rocks it likes into a digester that uses grinding, acids and catalysts.
It brings it back to a factory where the pellets of metal get melted down to ingots. Waldos that vaguely resemble human hands then conventionally process all the metal and carbon and so on the same way a factory works today. Just all the arms are vastly quicker and more certain, the quality control is much better, and the system runs 24/7, ceaselessly expanding on itself.
Or maybe the machine uses aneutronic fusion for power, giving it immense power density. It uses the far greater energy to blast all the rocks to plasma and separates them with magnets to pure elements. The pure elements are chemically reacted into gasses, and fed into a refrigerator sized brick of self replicating nanotechnology that copies both itself and also makes components for the larger machine.
Or maybe it’s some other way, some method using biology we know would probably work but it might suck in performance.
The point is, intelligence + materials + energy = self replication. There is no way you can be taken seriously if you don’t think a self replicating machine would be possible if you had AI. We already have self replicating machines, we just need human factory workers and engineers to perform critical steps the equipment cannot do for itself. And those humans are just another type of machine according to known laws of physics.
Chronos, did you know that most math problems that supercomputers are used to solve have series dependencies? Have you ever worked on an engineering team? If so, you’d know that there is something called a critical path. More than one person can work on that critical path, but each additional person adds overhead in that a person can individually think far faster than they can communicate with others.
There is a clear and enormous performance boost in solving problems if you have faster serial performance. Here’s a theorem explaining why.
If you’re going to postulate magic, I can’t really debate that. But that’s my point. Once you say the future will be magic, it stops all discussion about how we get there and what it will look like. Magic changes everything and in unforeseeable ways. The inability to lay out plausible steps to the future has been the death of Futurology since it cohered into a form of study in the 1960s. We obviously need to try to predict probably paths for the future, but before we get to everything changes magically we need to figure out what changes tomorrow and how to respond to it.
Can I ask why do you think a known fusion reaction that has been demonstrated in labs but can’t yet be harnessed (it requires too much confinement) is “magic”? Why is this a discussion ender? I simply mentioned it because it’s a logical top tier outcome given what we know will probably work.
Here are the fuels that will work : Aneutronic fusion - Wikipedia
I’m frankly a little annoyed with your characterization. If I said ‘zero point energy generator’, I’d agree, that would be magic. Or some method of harnessing a synthetic black hole. (though less magic, at least the equations balance for that way)
Or was it nanotechnology that I mentioned?
And how is it even relevant? Do you agree that you can build a self replicating machine without it, this would simply be the logical endpoint if you could solve the technical problems stopping you from doing it the best known way.
This article makes the assumption that each ‘age’ has to undergo a certain number of doublings before the next age can begin.
Is that assumption based on anything? I have no idea what the argument is. I don’t think scientists really know what started the agricultural revolution or the industrial revolution. Those topics are still being debated.
However most of the knowledge we have came after the industrial revolution. Each advance requires the advances before, and before the industrial revolution life didn’t change much. Going from 2300 BC to 2100 BC or from 300 AD to 500 AD were not that impressive. Going from 1780 to 1980 was a massive jump for human civilization.
Yes, of course faster processing would speed things up even further. I’ve already acknowledged that. I was replying to posts that seemed to be implying that that was the only possible way to ever speed things up, which is clearly false.