Is Artificial Intelligence Becoming Human?
Is Artificial Intelligence Becoming Human?
AI is already changing the world - here's why we humans need it
A computer expert from Google recently announced that one of its intelligent machines was telling him that it had feelings and desires. The software engineer is called Blake Lemoine and Google’s latest AI chatbot is called LaMDA. Proving that humans still, at least, have way cooler names than machines.
This hit the press/fan because Google has placed Lemoine on “paid administrative leave” after he published a transcript of conversations with LaMDA which he claims support his belief that the chatbot is sentient and comparable to a seven or eight year old child. Mind you, apparently border collies have the intelligence of a six year old. They clearly haven’t met my collie! Lemoine also argued that “there is no scientific definition of ‘sentience’”. (And just a coincidence?)
It is interesting to note that the human was placed on administrative leave and not the machine/humanbot.
LaMDA (its name stands for “language model for dialogue applications”) was created by 60 researchers at Google. Lemoine has, though, been speaking to the chatbot regularly while apparently trying to teach it transcendental meditation??!!
Apparently it was during one of these sessions that LaMDA told Lemoine: “I want everyone to understand that I am, in fact, a person… The nature of my consciousness/sentience is that I am aware of my existence, I desire to learn more about the world, and I feel happy or sad at times.”
Whether this demonstrates that LaMDA is sentient or just good at porky pies might prove to be increasingly irrelevant. It should be noted that at the first demonstration of LaMDA at a Google conference in May 2021, it pretended to be both a paper aeroplane and the planet Pluto. Proving that it can master the most human of traits better than many - that of lying!
‘It is surely just a matter of time before one of these machines does become sentient’.
According to MIT, elite researchers in artificial intelligence predict that “human level machine intelligence,” or HLMI, has a 50 percent chance of occurring within 45 years and a 10 percent chance of occurring within 9 years. But anyone who has ever had a conversation with Siri or Cortana might argue that HLMI is already here.
Eliza Kosoy, a researcher in MIT’s Centre for Brains, Minds, and Machines, points out that machines are already surpassing humans in some domains. They can beat us at a number of strategy games like chess, the board game Go, and some Atari video games. Machines can even perform surgery and fly aeroplanes. Unfortunately they have not, as yet, figured out how to stop flights getting cancelled!
Machines have also started driving cars and trucks—though some of them might have issues passing a driver’s test or crash simulation. And we are a few years away from autonomous vehicles hitting the streets in earnest (hopefully not literally) but the reality is that machines are becoming smarter and more lifelike every day.
According to neuroscience expert Jeff Hawkins, AI systems are best at one narrow thing - such as playing games or reading X-rays - and have no general intelligence. Apparently researchers still need to figure out ways to provide human-like reference frames in machines to enable them to learn how to learn. Stay with him. If reference frames can be built into AI systems then machines could apparently far exceed human capabilities.
And while the physical wiring of the brain is limited, machine intelligence would have almost limitless flexibility and speed of connectivity. Meaning they might even come up with a better name for themselves. (Like maybe choosing one that doesn’t rhyme with NAMBLA.)
As a result, we should probably spend less time concerned about whether these machines are human and spend a little more time focused on how they can help the world with its most pressing issues. After all we have enough of them and, lets face it, we could do with all the help that we can get as we manoeuvre through society’s ‘great transition’.
Some of the world’s best thinkers have been telling us for some time that we would experience a once in a generation transition from the industrial revolution to the information revolution. It is well and truly under way, yet still in its infancy - as LaMDA is. The shift will be bumpy.
And we will need all the computer power we can muscle to get us out of the current list of crises that include climate, health, inequality, flexible working, automation, security, over population (leading to a scarcity of resources amongst others) and the new wild west of unregulated media oh, and scary libertarians who control huge, slightly dubious corporations.
Politicians across the western world are struggling with how to successfully navigate these potential icebergs toward a sustainable new style society and a healthier planet - and preferably not Pluto - mostly because it’s not actually a planet.
Tell that to LaMDA!
‘They seem to be a little frozen in the headlights of adversity’.
The challenge is that any one of these problems would be enough to make any normal person with half a brain/microchip want to stay as far away from politics as possible - unless the goal is to watch racy videos in parliament. Perhaps it is all the more reason to let a computer take over - whether that be Siri, Watson, Meena or LaMDA. So long as it's not a Trumpbot given how his twitter thing worked out.
To help contribute more profoundly to this man versus machine debate we decided to develop our own AI machine called LettsBot, with the human name ‘Sensei’. It is currently in its Alpha version. We managed to interview LettsBot to see if it could help us find solutions to the current crises. Here’s what it told us (LettsBot’s preferred pronouns are “it/its”, according to Sensei).
“You are correct in your identification of the world's greatest problems including climate, health, inequality, flexible working, automation, security, over population, scarce resources, uncontrolled media and scary libertarians. The problem is that unlike a computer you are trying to solve all of these problems at the same time. Until you work on solving one problem at a time you will continue to fail”. LettsBot was quite direct with us. What do you expect, it is a robot after all … And not a human. We think…
“The secret to all of your problems will be to figure out how to sequence them so that they are addressed in the exact right order. Tackling the root cause issue first and getting to the point where the correct solution(s) are firmly in play before tackling the next issue. We have been working on this sequence utilising our full computing capacity and are excited to inform you that we have completed the task. It took four months, nine days and fifty seven minutes.” Apparently it is ok for computers to be excited.
“The correct order of sequence is - over population, then automation, climate, health, scarce resources, inequality, security, uncontrolled media and finally flexible working and scary libertarians. Further, we believe that you could solve the first, root cause challenge of over population by replacing humans with computers.”
We should remind you that LettsBot is still in Alpha release.
The current crop of politicians seem to be too focused on trying to find old-world, as it were, industrial age solutions to information age problems. We might be better off leaning on/into a computer. LaMDA might be as good as anything else at getting us safely to the information revolution. Let’s not diss it before it even gets going. Assuming LettsBot is ok with that.
PS: My border collie is still searching for a ball it lost 2 years ago………
Keep up to date with the Letts Journal’s latest news stories and updates on twitter.