Yes – because you are human, you can ascertain what the tone of this essay will be, along with the subject.
No – there is no computer that can both determine the subject (Transhumanism-Robots-AI-Hopium) and realize the tone and direction we will be taking here in this essay.
Why? Well, computers, even the much-vaunted artificially intelligent ones, are PDD (pretty damned dumb). I use the above part of this essay as proof of that very thing. Let’s call the ‘leading’ title of this essay exhibit “A”, because an AI computer cannot follow the slim data your brain can and arrive at a general direction of thought. In fact, an AI will not know what to make of the term “leading title”, other than l’leading’ means it precedes another object. It will not realize the other, more humanistic use of the word “leading”.
So let’s just sort of lay out some rudimentary essay research here:
ONE second of thinking like a human takes nearly 83,000 processors? Seems the brain is kind of complex then, compared to supercomputers. Some additional research also tells us that the computer this simulation was run on is called the “K” computer. It uses 9.89 megawatts of power to operate, the equivalent of 10,000 suburban homes. Now that’s really great, but I can ride my bicycle in traffic all morning after eating 2 eggs, 2 strips of bacon and some juice. Hmmmm….
Let’s call that exhibit “B” – supercomputers are terribly huge, expensive, inefficient and immobile. Yet they are the only type of computer that can mimic a single second of human neural activity.
One of the biggest issues with human+computer thinking is that the brain is analog – it carries an immense amount of information that is scalar in nature, because the signals are sent and received with variable frequencies and amplitudes. Digital systems are purely frequency based – a bit is either on or it is off. In an analog system, a bit can be in on, off, standby, overdrive and all points in between. Our analog data is communicated by a network of massively interconnected electro-chemical cells, some of which are highly specialized and many are more general in nature.
If we take exhibits A + B, we can extrapolate that to mimic the brain of a mouse for a second would take a computer the size of a house, if only we could miniaturize what we have today, reduce the energy requirements by a mere 10,000% and have a better way to interconnect the CPU modules. In the future, if we get even more miniaturization (getting harder every year) and more efficient interconnection (again, very hard to get much more than a 3D array), we could do what a mouse does for an entire second of time, but the power required is still much more than a house, because there are limits to silicon technology.
I agree that the Terminator movies and the Star Wars movies were great, but we are nowhere near that, not even in decades. Even with quantum computing, there are power issues, density issues, storage issues and giant manufacturing issues. The power for the Terminator was a nuclear battery, and nobody even has a clue how one might work. Shielding it alone would be a nightmare, and the politics of it? The takeaway here is that until a supreme amount of innovation happens in both the CPU engineering and in the power consumption curves, there are some very big hurdles to having a computer remotely resembling mouse intelligence that isn’t as big as a navy destroyer. A generational timeline is more appropriate when you look at the pace of scientific advance coupled with engineering advances.
Let’s look at what some of the current crop of tekkies had to say recently:
Some object-recognition software today “is 90% accurate, you go to Facebook, there are just so many faces — [but there is] 90% accuracy” in identification. Still, even at 90% “your computer-vision colleagues would tell you ‘that’s dumb’…. But to get from 90% accuracy to 99% accuracy requires a lot more data” — exponentially more data. And then to get from 99% accuracy to 99.9% accuracy, guess what? That needs even more data…….“And that’s why I think self-driving cars, which involve taking actions based on data, are extremely hard [to perfect]…. “Yes, it’s a great concept, and yes, we’re making major strides, but … to solve it to the point that we feel absolutely comfortable — it will take a long time.”
“Here’s a simple fact. To fly a drone requires about 200 watts per kilo. So, if you want to lift a 75-kilo individual into the air, that’s a lot of power. Where are you going to get the batteries to do that?” The only power source with enough “power density” to lift such heavy payloads is fossil fuels. “You could get small jet turbines to power drones. But to have electric power and motors and batteries to power drones that can lift people in the air — I think this is a pipe dream.”
“You look at electric motors, it’s World War II technology. So, on the physical side we are not making the same progress we are on the information side. And guess what? In the U.S., 2% of all of electricity consumption is through data centers. If you really want that much more data, if you want to confront the hockey stick, you are going to burn a lot of power just getting the data centers to work. I think at some point it gets harder and harder and harder….”
“… A deep-learning algorithm that can do just speech recognition, which is translating what you are saying, has to be trained on millions of hours of data “and uses huge data farms,” Fung noted. And while a deep-learning network might have hundreds of thousands of neurons, the human brain has trillions. Humans, for the time being, are much more energy-efficient. They can work “all day on a tiny slice of pizza,” she joked.
These are some very bright, highly motivated engineers we are quoting here. They are working on the cutting edge of AI and robotics. They are in agreement that the evolution the media is hyping is generations away. They are also in agreement that powering all of this is not going to work with batteries, as we know them today.
Here is another article that lends some credence to the direction of this essay:
Kennedy is a true believer – his actions indicate that, whether you believe them heroic or foolhardy. What is most interesting here is that he glossed over several issues that will always plague any interfacing with the brain. One is that the brain floats in a bath of cerebral fluid and moves. That is what absorbs shocks and keeps this tender neural organ safe and happy in our skulls. Tying an electrode or a golden net into the brain, where one end of it is anchored in neural tissue and the other must exit the skull via a plug – problematic for the potential to damage the brain from impact.
The healing is also problematic. There are many instances where attaching electrodes produces scarring in the nearby neurons or accretions of unwanted tissues. There are people whose bodies always reject piercings, and the same is true for brains – some bodies will not tolerate a foreign object and actively work to expel them. Some bodies treat any foreign object as an infection, which makes rejection very severe.
Finally, there is Kennedy’s work, which is still in limbo. Deciphering the way the brain talks within itself and handles data to form knowledge – that is very honestly generations away. And yet there are those who believe we can mimic the human brain, when we do not even know how it moves and resolves data into knowledge. In other words, nobody knows how I am able to think; to collate facts, weight them, infer other lines of thought, review experiences and draw a conclusion. If we cannot describe how the brain functions effectively, then how in hell do these people think we are going to make an artificial intelligence that actually comes close to a human being? We have yet to define what makes a human, well… human.
Another few fun facts to digest revolve around (now legal) bionic eyes. But we need to know what the eye has in terms of bandwidth; the human eye sees in near-limitless colors and has an optic nerve attached to each of your retinas. The diameter of the optic nerve is around 3mm, and each contains around 1,400,000 nerves for each eye.
The bionic eyes they are currently implanting contain 40-150 electrodes (nerve equivalents) that can transmit digital (on/off – black/white) signals. There are many problems, the largest of which is that these electrodes stimulate groups of nerves due to their size relative to the actual individual optic nerves. This causes ‘streaking’ and ‘meteors’ per the patients who have implants, likely caused by their own nerves “shorting out” via the larger implanted electrodes.
Bio-engineers are hoping that someday, using nanotechnology, they can produce an implant that contains 1096 electrodes. <– read that again We have, within each eye, 1,400,000 nerves from nature that see in all colors, and we are hoping to, some day, have 1096 that see in black and white.
The final subject is robots, which are coming to take our jobs, one and all – not. There was a company named Boston Dynamics that was making robots for military use using DARPA funding. If you search for this company, you will find some quite scary looking robots. It is easy to imagine them armed with lasers or mini-guns and just killing people like cutting wheat in a battlefield – they have no fear or remorse due to being robots.
In some of the videos, you can hear what sounds like a leaf blower running in the background. It actually is a leaf blower motor, running the generator to power the robot. I refer you back to the quotes from the tekkies above, in particular regarding the subject of drones carrying a person. I ask you to remember the nuclear battery that the Terminator robot had, because we don’t have anything remotely like that. Hence these larger robots require an ICE (internal combustion engine) and generator to do any serious amounts of work.
Think about your rechargeable electric drill or electric weed whacker. You have about 15-30 minutes continuous use, under normal load, if you have a fully charged, large battery. Then you have to swap a new one in. Now imagine running the actuators, servos, sensors, step motors and the computers and software for one of these robots on lithium batteries. In the big ones, they don’t even try – they have an ICE/generator within the chassis.
This was enough to put the army off the trail for everything except the self-balancing and mobility of these bots. Hard to hide the IR signature or ignore the whine of the motor on the big ones. And yet the army needs the big ones, because they want to replace humans.
So everything is waiting for the “Terminator battery” in robotics. A Japanese company recently bought Boston Dynamics, which is not surprising since the Japanese have adored robots for decades. Don’t believe me – go watch some anime.
This is also why you are seeing the army and DARPA latching onto the smaller robots being used as a swarm. This is actually doable, but thinking they will survive an EM battlefield is questionable at best. If DARPA wants them to be smart and autonomous, then we are back at the AI issue, which we have already addressed. What is more likely is a highly programmed bot running an expert software routine specifically designed to differentiate between IR signatures and IFF signals. That way only the bad guys get targeted, but it also means that good guys must have an IFF on their person. Electronic warfare with bots will be bloody and very unforgiving.
Robots have the same problem as drones – power density required to perform serious work. Until either battery tech or electric motor tech take a quantum leap, bots that can do serious work will have a plug or an ICE.
I do not want to discourage anyone from pursuing work in any of these fields. My vision always was of benevolent robots doing things that break peoples backs – like pulling weeds and clearing fence lines. I think robots have quite a future ahead, as does AI. Transhumanism? Well, let me just say that Kennedy was the first and won’t be the last. Yet until some of these many issues are worked out, interfacing man and machine will likely end in train wrecks or aborted projects.
I do want people to read and think about where we are, what the real issues are, and be able to discern doom porn and hype from media outlets looking for clicks and viewers. There aren’t too many lies regarding these fields, but there is quite a bit of omission and literally no backstory out there from people in these industries. In some cases, companies are seeking financing, and they will be very “forward looking” in what they release to the public. When the Boston Dynamics videos were first released, the volume was either turned down tremendously or the videos were mute. To think this wasn’t intentional would be foolish. Similar things happen when companies try to put the best face they can, on what they have, to generate interest
It’s interesting to see that the tekkies think it is mid-level workers most at risk of AI replacing them, isn’t it? And I am very happy that they are seeking creative people rather than just those with multiple degrees or good GPA’s. I think there is a good future for those wishing to work in robotics, even if we don’t get the Terminator battery. I think AI is promising too, and I don’t worry about an AI “fooling” me on the phone – they are easy to trip up and I don’t live on my phone. AI running the stock market has train wreck written all over it. AI will likely hit those jobs where process is the most important – where creativity is not always desired. Authoring fiction? I think it will be quite a while before AI rivals a good author with command of characters, plot and the native language. Maybe comic books…?
There is little more human than creativity, and it will be needed to handle the many issues these technologies bring as they mature with our children, and their own.