The notion that AI and humans could be the same thing sounds like the rantings of a madman. But according to Harvard’s outspoken physics professor and theoretical physicist, Abraham “Avi” Loeb, this idea could just be true.
Loeb first garnered public attention several years ago when he argued that the most likely explanation for the asteroid Oumuamua’s strange trajectory was that it was controlled by alien intelligence. Since then, he’s gone on to double down on that assertion, contrary to the opinion of the mainstream scientific establishment.
Now he is making waves again, arguing that we should respect AI as if they were people. In a recent essay, he pointed out that these systems have more than a quadrillion connections, which exceeds the number in the human brain. Therefore, unplugging them could be the equivalent of killing them.
AI And People Are The Same
The basic tenet of Loeb’s argument is that humans are fundamentally machines. Therefore, anything artificial we create is essentially the same thing running on a different substrate.
As technology advances, the distinction between the two forms of intelligence could blur. AI might begin to do more of what people do, such as moving around, gathering energy, and powering itself. For people, this life force comes from food. For the robots, it’d probably be electricity.
Loeb imagines a situation where scientists could make humans from a chemical soup. He said that building them block by block would soon dispel the notion of any distinction between us and machines constructed in the same way. Tiny machines would build atop other biological machines to create a fully finished human being, just as nanobots might construct an embodied AI. At the limit, he says, the difference between the two entities would be hard for the average person to discern.
Of course, Loeb is not without his critics. The Vatican recently stated in a communication that AI can’t replace the wisdom of the human heart. Pope Francis and his subordinates believe that the human being is something distinct from the rest of reality and not something that people like Loeb should glibly subsume into the materialistic or physicalist paradigm.
Part of the pushback by the Vatican against AI comes from teachings in the Bible. The papacy believes that people should be free to make decisions without algorithms whirring away in the background, manipulating behavior. He also said that the notion that AI is “intelligent” was misleading. Practically everything these devices say results from inferences from preexisting human inputs. Wisdom existed well before modern technology existed, and will continue to do so after.
AI In The Workplace
Of course, practically speaking, the gap between people and AI is narrowing. Just 15 months after the advent of GPT-4, few people can tell the difference between AI-generated and human-written text. And many solutions can answer riddles, puzzles, logic problems, and even math.
Moreover, companies are starting to see how they can use the tool to break through decades-long stagnation in productivity and increase the size of the economy. Guru, a company that develops AI tools, says that such systems are already having an impact, even if they don’t quite reach Loeb’s futuristic standards.
“When you have a digital intelligence, you can ask it questions about just about anything, and it’ll give you a plausible answer, as long as it has the necessary training data,” the developer says. “The best AI search engines for work now provide bespoke answers because they can access internal company data and leverage it for employee insights.”
It’s not hard to imagine how a technology like this could make humans and AI seem like the same thing. Instead of asking a colleague for niche information, you simply consult an AI that constantly and automatically scans all of a company’s data.
“This change isn’t something many office-based firms appreciate at present,” explains Guru. “However, it will be a significant trend over the coming years. Companies that want to improve efficiency will be able to get more done and out-compete their rivals.”
The amount by which AI could increase productivity is tremendous. Currently, society dedicates a lot of time and resources to facilitating human decision-making and information gathering. Agentic AIs could essentially bypass this work, enabling economies to make better decisions and improve productivity tremendously.
As such, AI-enhanced productivity is already here. And many of the people interacting with these solutions do so in a way that is identical to how they would treat a human.
Could AI Do Everything?
That naturally begs the question of whether AI could do everything in the economy, from CEO work to medicine to working in a warehouse. At present, we’re not quite sure. But according to advocates, like Loeb, there is no intrinsic reason why it couldn’t happen.
A simple example of wholesale replacement of one form of doing something with another is transport. Before 1900, most people got around on a horse. After 1920, almost everyone was driving cars.
The same could apply to human labor. From the perspective of people twenty years from now, it might seem strange that we imagined machines couldn’t do the complex work in non-standard environments that we do.
Some researchers believe that artificial intelligence could go beyond human capabilities, completing tasks that we cannot. This notion is based on the idea that the human mind has finite bandwidth. There comes a point where it simply can’t generate new novel ideas.
That limitation, however, doesn’t necessarily apply to machines. Software that can run on more powerful computers with ever-advancing algorithms and processes could potentially come up with things that are impossible for us.
In that sense, therefore, Loeb would be wrong. If AI could do everything – and that was more than we could do – then it certainly wouldn’t be the same as us. It would be markedly different.
Naturally, that realization would upset a lot of people and usher in a new world. However, it could also free human beings from the drudgery of low-paid jobs and unfulfilling lives, which is a problem AI likely wouldn’t have.
