A Brief History of Human Technology

Before I write about artificial intelligence and its potentially pivotal role in history, I want to provide a brief history of human technology. As I noted in my last post in this series, human beings don’t and possibly can’t live without any technology at all. But, for most of our history, these technologies have been relatively simple, at least from our modern perspective. To get a better understanding of how dramatically and rapidly our technologies have changed, let’s consider some timelines.

Millions of Years of Basic Tool Usage

In a very real sense, we humans have long been expanding our capabilities via technologies for hundreds of thousands of years. In fact, we were likely doing it long before we were even humans. Today, there are examples of tool usage among all the non-human great apes (bonobos, chimpanzees, gorillas, orangutans, and human), which probably means that our common ancestors were also users of tools.

Consider the Hominin timeline below, for example. Our ancestors split aways from the ancestors of today’s chimpanzees about 8 and half million years ago, and there’s a good chance those ancestors used wooden tools in ways similar to today’s chimps and bonobos. They do things such as use sticks to fish termites out of mounds and to dig for tubers, wield stones to crack nuts, and even employ leaves to soak up water or wipe their mouths.

From Wikipedia with small modifications by me

A Timeline of Inventions and Technological Advances

As timeline above shows, however, a rapid growth of tools and technologies began with the advent of Homo sapiens. Although the flowering of various technologies arose with homo sapiens over a period of tens of thousands of years, there was a massive uptick in new and powerful technologies at around the start of the Industrial Revolution. Consider the following list of some of the most important inventions, though obviously many of these dates are, at best, estimates:

900,000 years ago: Hafting
400,000 years ago: Spears
200,000 years ago: Language
170,000 years ago: Clothing
100,000 years ago: Boats
90,000 years ago: Harpoons
70,000 years ago: Arrows
47,000 years ago: Mining
42,000 years ago: Tally stick
36,000 years ago: Weaving
28,000 years ago: Ceramics
28,000 years ago: Rope
23,000 years ago: Domestication of dogs
16,000 years ago: Pottery
12,000 years ago: Agriculture
9,000 years ago: Alcohol
8,000 years ago: Irrigation
7,000 years ago: Copper smelting
6,000 years ago: Plumbing
6,500 years ago: Lead smelting
5,500 years ago: Domestication of horse
5,300 years ago: Written word
4,300 years ago: Abacus
4,200 year ago: Protractor
3,500 years ago: Glass
3,300 years ago: Water wheel
3,300 years ago: Iron smelting
2,650 years ago: Crossbow
2,650 years ago: Windmill
2,485 years ago: Catapult
2,200 years ago: Paper
1,803 years ago (220 AD): Woodblock printing
1,573 years ago (450 AD): Horse collar
1,446 years ago (577 AD): Sulfur matches
1,405 years ago (618 AD): Bank note
1,223 years ago (800 AD): Gunpower
935 years ago (1088 AD): Movable type
695 years ago (1326 AD): Cannon
584 years ago (1439 AD): Printing press
525 years ago (1498 AD): Rifle
418 years ago (1605 AD): Newspaper
415 ( years ago 1608 AD): Telescope
403 years ago (1620 AD): Compound microscope
393 years ago (1630 AD): Slide rule
381 years ago (1642 AD): Mechanical calculator
367 years ago (1656 AD): Pendulum clock
343 years ago (1680 AD): Piston engine

Start of the Industrial Revolution

290 years ago (1733 AD): Flying shuttle
259 years ago (1764 AD): Spinning jenny
258 years ago (1765 AD): Steam engine
230 years ago (1793 AD): Cotton gin
219 years ago (1804 AD): Railway
216 years ago (1807 AD): Steamboat
197 years ago (1826 AD): Photography
195 years ago: (1828 AD): Reaping machine
179 years ago (1844 AD): Telegraph
147 years ago (1876 AD): Telephone
147 years ago (1876 AD): Internal-combustion engine
144 years ago (1879 AD): Electric light
138 years ago (1885 AD): Automobile
122 years ago (1901 AD): Radio
120 years ago (1903 AD): Airplane
97 years ago (1926 AD): Rocketry
96 years ago (1927 AD): Television
86 years ago (1937 AD): Computer
81 years ago (1942 AD): Nuclear power
76 years ago (1947 AD): Transistor
72 years ago (1951 AD): First artificial neural network
70 years ago (1953 AD): Structure of DNA discovered
68 years ago (1955 AD): Artificial intelligence term coined
66 years ago (1957 AD): Spaceflight
65 years ago (1958 AD): Perceptron, artificial neural network for pattern recognition
64 years ago (1959 AD): Machine learning term coined
50 years ago (1973 AD): Cell phone
49 years ago (1974 AD): Personal computer
49 years ago (1974 AD): Internet
39 years ago (1984 AD): 3D-printing
28 years ago (1995 AD): DNA sequencing
11 years ago (2012 AD): CRISPR
8 years ago (2014 AD): Generative adversarial network AIs
5 years ago (2018 AD): Generative pre-trained transformer AIs

These technologies are all now part of the our technosphere. If we picture that sphere as a kind of balloon, then we can see that it filled up relatively slowly at first but picked up momentum around 40,000 years ago, then really took off about 400 years ago.

Are Breakthroughs Speeding Up or Slowing Down?

The Speeding Up Theory

Some thinkers believe that we are in the midst of a virtual explosion of technology. Futurist Ray Kurzweil claims that we are in a state of exponential technological growth driven by the law of accelerating returns.

Back in 2021, he wrote, “An analysis of the history of technology shows that technological change is exponential, contrary to the common-sense ‘intuitive linear’ view. So we won’t experience 100 years of progress in the 21st century — it will be more like 20,000 years of progress (at today’s rate). The ‘returns,’ such as chip speed and cost-effectiveness, also increase exponentially. There’s even exponential growth in the rate of exponential growth. Within a few decades, machine intelligence will surpass human intelligence, leading to The Singularity — technological change so rapid and profound it represents a rupture in the fabric of human history. The implications include the merger of biological and nonbiological intelligence, immortal software-based humans, and ultra-high levels of intelligence that expand outward in the universe at the speed of light.”

Well, wow, that’s a lot. It is the epitome of techno-optimism (if merging with machines is your idea of optimism). On the other side of the coin, of course, are those who are quite certain that superintelligent AI will mean the end of humanity.

But I think the primary difference between techno-optimism and techno-pessimism boils down to one thing: AI’s future role in the creation of the technosphere. We’ll get to that in the next post. In the meantime, however, let’s consider the idea that technological change is actually slowing down.

The Slowing Down Theory

Certainly, on a geological time scale, all these inventions we’ve listed have arisen virtually simultaneously. But we don’t live in geological time and some experts believe that, from a human point of view, there’s been a dramatic slowdown in true innovation and scientific breakthroughs in recent years.

The authors of the study titled “Papers and patents are becoming less disruptive over time” analyzed data from 45 million papers and 3.9 million patents across six decades (1945–2010). Tracking how their disruption index changes over that timeframe, the researchers found papers and patents are increasingly less likely to be disruptive.

For example, in the area of patents, the decline in disruptiveness between 1980 and 2010 ranged from 78.7% for computers and communications to 91.5% for drugs and medical. They write, “Our analyses show that this trend is unlikely to be driven by changes in citation practices or the quality of published work. Rather, the decline represents a substantive shift in science and technology, one that reinforces concerns about slowing innovative activity. We attribute this trend in part to scientists’ and inventors’ reliance on a narrower set of existing knowledge.”

So, what can we do differently to address this issue? The authors suggest:

To promote disruptive science and technology, scholars may be encouraged to read widely and given time to keep up with the rapidly expanding knowledge frontier. Universities may forgo the focus on quantity, and more strongly reward research quality, and perhaps more fully subsidize year-long sabbaticals. Federal agencies may invest in the riskier and longer-term individual awards that support careers and not simply specific projects, giving scholars the gift of time needed to step outside the fray, inoculate themselves from the publish or perish culture, and produce truly consequential work.

The Extension of the Human Mind

Whether the creation of disruptive technologies and scientific paradigms is speeding up or slowing down, it’s clear that we have recently made large breakthroughs in artificial intelligence, which is an extension of our cognitive capabilities

Of course, we humans have been aiding and extending our mental capacities at least since the tally stick and probably long before then. Books, photos, maps, calculators, spreadsheets, word processors and much more have all been extensions of our minds.

But generative AI does feel like a much further extension, capable of doing various things that only the most capable and educated of people could have done before now. For example:

ChatGPT’s Performance on Academic and Intelligence Tests
The Uniform Bar ExamWhile GPT-3.5, which powers ChatGPT, only scored in the 10th percentile of the bar exam, GPT-4 scored in the 90th percentile with a score of 298 out of 400
The SATGPT-4 aced the SAT Reading & Writing section with a score of 710 out of 800, which puts it in the 93rd percentile of test-takers
The GREWhile it scored in the 99th percentile on the verbal section of the exam and in the 80th percentile of the quantitative section of the exam, GPT-4 only scored in the 54th percentile of the writing test
USA Biology Olympiad Semifinal ExamGPT-4 scored in the 99th to 100th percentile on the 2020 Semifinal Exam
AP ExamsGPT-4 received a 5 on AP (advance placement) Art History, AP Biology, AP Environmental Science, AP Macroeconomics, AP Microeconomics, AP Psychology, AP Statistics, AP US Government and AP US History. On AP Physics 2, AP Calculus BC, AP Chemistry and AP World History, GPT-4 received a 4
IQEstimated on the basis of five subtests, the Verbal IQ of ChatGPT was 155, superior to 99.9 percent of the test takers. It was not able to take the nonverbal subtests. But the Verbal IQ and Full Scale IQ scales are highly correlated in the standardization sample.

Although this table highlights the power of these technologies, it leaves aside their lack of “common sense,” their poor mathematics capabilities (for now) and their chronic habit of hallucination and confabulation. These and other issues are why few view these technologies as actual artificial general intelligence.

But this doesn’t mean that such AI doesn’t already play a fast-evolving, uniquely creative and increasingly pivotal role in the shaping of our technosphere. That will the subject of my next post.

The Singularity Just Got Nearer…Again…And How!

To me, it already it seems like another era. Last October, I wrote a tongue-in-cheeky post called “The Singularity Is Pretty Damned Close…Isn’t It?” I wrote it after the AI art generator revolution had started but before ChatGPT was opened to the public on November 30, 2022. That was only four months ago, of course, but it feels as if everything has sped up since, as if we human beings are now living in dog years. So it’s already high time to revisit the singularity idea.

Are We Living on Hertzian Time Now?

As you may know, the word “hertz” — named after Heinrich Rudolf Hertz, a German physicist who discovered electromagnetic waves in the late 19th century — is a unit of frequency. More specifically, it’s the rate at which something happens repeatedly over a single second. So, 1 hertz means that something happens just once per second whereas a 100 hertz (or Hz) means it’s happening 100 times per second.

So, an analog clock (yes, I still have one of those) ticks at 1 Hz.

 Animation of wave functions, by Superborsuk

Unless you’re an engineer, you probably think about hertz as part of the lingo folks throw around when buying computers. It’s basically the speed at which central processing units do their thing. So, a laptop with a speed of 2.2 GHz has a CPU that processes at 2.2 billion cycles per second. Basically, that’s the speed at which computers carry out their instructions.

So, my (completely fabricated) notion of Hertzian time refers to the fact that, day to day, we humans are seeing a whole lot more technological change cycles (at least in terms of AI) packed into every second. Therefore, four months now feels like, well, a whole lot of cycles whipping by at a Hertzian tempo. Generative AI is overclocking us.

How Wrong Can I Get?

Back in late October, I wrote, “There’s at least one pursuit that AI has yet to master: the gentle art of conversation. That may be the truest assessment of human level intelligence. At least, that’s the premise underlying the Turing test.”

Many Hertzian cycles later, the world looks very different. Now millions of people are chatting up these proliferating LLMs (I just got my access to Bard the other day, btw) every moment of every day, and we’re just getting started.

It’s true that if you get used to conversing with these models, you can tell that they aren’t quite human. And, the main ones go to some length to explain to you, insist even, that they are NOT human.

Everyday Feels A Little More Turing Testy

I recently specifically asked ChatGPT3, “Do you think you could pass the Turing Test if properly prepared?” and it responded: “In theory, it is possible that I could be programmed to pass the Turing Test if I were given access to a sufficiently large and diverse dataset of human language and provided with sophisticated natural language processing algorithms.”

I tend to agree. The newest AIs are getting close at this stage, and I imagine that with only a few modifications, they could now fool a lot of people, especially those unfamiliar with their various little “tells.”

Coming to Rants and Reality Shows Near You

I think society will increasingly get Turing testy about this, as people debate whether or not the AIs have crossed that threshhold. Or whether they should cross it. Or whether AIs have a soul if they do.

It’ll get weird(er). It’s easy imagine growing numbers religious fundamentalists of all types who demand Turing-level AIs who preach their particular doctrines. And who deem those “other” AIs as downright satanic.

Or envision reality TV shows determined to exploit the Turing tests. Two dozen attractive, nubile wannabe LA actors who are trying to out-Turing one another on a tropical island. They win a cool mill if they can tell the (somehow telegenic) AI from the (oh-so-hot) real person on the other side of that sexy, synthesized voice. Think of the ratings!

Kurzweil May Have Nailed It

As I said in that first singularity piece, the futurist Ray Kurzweil has predicted that an AI will pass the Turing Test in 2029. I wasn’t so sure. Now I wonder if it won’t be sooner. (I suspect the answer will depend on the test and expertise the people involved.)

But will the passing of the Turing Test mean we are right smack in the middle of the singularity? Kurweil doesn’t think so. He has his sights set on 2045 when, as I understand it, he thinks humanity (or some portion of it) will merge with the superintelligent AIs.

That still seems very science fictional to me, but then I also feel as if we’re all living right smack dab in a science fictional universe right now, one I never thought I’d live to see….

Those Seas Are Rising Fast

My predictions on the rising seas of AI generated media, however, are still looking pretty good. Of course, I’m not alone in that. A 2022 Europool report noted, “Experts estimate that as much as 90% of online content may be synthetically generated by 2026.”

What’s going to make that number tricky to confirm is that most media won’t be fish or foul. It’ll produced by a combination of humans and AIs. In fact, many of the graphics in my blog posts, including this one, are already born of generative AI (though I try to use it ethically).

Are These the Seas of the Singularity?

The real question to ask now is, “Are we already in the singularity?”

If we use the metaphor of a black hole (the most famous of all singularities), maybe we’ve already passed the proverbial event horizon. We’ve moved into Hertzian time and overclocking because we’re being sucked in. From here, maybe things go faster and faster until every day seems packed with a what used to be a decade’s worth of advances.

These rising seas, the virtual tsnamis, might just be symptoms of the immense gravitational forces exerted by the singularity.

Or maybe not….Maybe such half-baked mixed metaphors that are just another sign of West Coast hyperbole, bound to go as disappointly bust as the Silicon Valley Bank.

Time’ll tell, I guess.

Though it’ll be interesting to find out if it’s normal time or the Hertzian variety.