Recently, Sina Finance had a conversation with Matthew Cobb, a professor of zoology and neuroscientist at the University of Manchester in the UK. He is one of the few contemporary neuroscientists who possess both scientific insight and philosophical

In this exclusive conversation spanning 37 minutes, Cobb talks about the metaphors of science history, the illusions of artificial intelligence, the significance of creativity, and his sincere advice for young generations.

As a biologist and an "AI skeptic," Matthew Cobb said that so far, he has not seen any AI system produce anything truly creative.

AI and the human brain are two completely different systems

Matthew Cobb said that when writing The Brain as Computer, he discovered a very interesting historical fact. Today we often say that the brain is like a computer, but actually, computers were initially designed based on the human brain. The foundatio

When digital computers were invented in 1945, their designers chose to use a digital logic structure because of an influential paper that suggested neurons in the human brain are connected in a logical circuit fashion. Based on this biological hypoth

That is why all computers today are based on that system.

However, people soon found that the human brain actually doesn't work in that way. The brain isn't a digital system with logical states of 'on' and 'off'. It's neither linear nor stable - because it's alive and evolved.

So, initially it was believed that 'computers are like brains'; soon, this belief was reversed into 'brains are like computers'.

In other words, the brain and the computer are two different systems: one is a biological system that has evolved over billions of years as a product of natural selection, while the other is an artificial system designed by humans in just seventy yea

AI will change the way of data processing, but cannot unlock the mysteries of the brain

Matthew Cobb also said that AI will indeed change the way neuroscientists process data.

In the field of neuroscience in the 21st century, a major challenge lies in the fact that we have an abundance of data but are unable to interpret it. We have vast amounts of data on cell types, molecular biology, and cellular electrical activity, bu

AI can truly detect patterns that are invisible to the human eye and statistical models.

So I believe that in the future, we will deliver a large amount of neuroscience data to AI, allowing it to help us identify potential patterns and find connections that we have not noticed. In this regard, AI will become a valuable tool.

Even so, he still doesn't believe that AI can become a 'model' for explaining the human brain, or even the brain of a simple fruit fly larva.

The mechanism of AI systems is far beyond our cognitive scope - current theories cannot correspond to it, and computers can never become a magic stick to solve the mysteries of neuroscience for us.

I am a biologist. And biology - is always full of wonders and complexities.

The human brain is the result of evolution, but we still don't know how it happened.

I asked Cobb if he was certain that our brains weren't "designed".

Matthew Cobb replied, 'Yes, I am very certain.'

The brain is not a product of design. Everything in biology shows its 'chaotic' nature - a result of billions of years of evolution. We can clearly see the imprint of this evolution in the way the brain works and in the genes involved in its formatio

We can even trace the origin of the brain, which dates back to the emergence of animals, more specifically, when animals first exhibited bilateral symmetry, around 700 million to 1 billion years ago.

At that time, a group of nerve cells emerged that can sense the external environment. Imagine a primitive worm: it needs to know whether food is on the left or right, thus developing cells that can sense direction. Consequently, it has a front end fo

After billions of years of evolution, that front end slowly turned into a brain - ultimately, our ability to think and be aware is an astonishing miracle in itself.

But we still have no idea how it happened.

AI systems do not possess true "creativity"

As an AI skeptic, Matthew Cobb said that he hasn't seen any AI system produce something truly creative yet.

People give AI a prompt, and it generates images or videos. For example, putting Marilyn Monroe in a modern TV show - this is not creation, just a humorous mashup.

Computers have not yet created truly "new" things. The reason is quite simple - they are based on what has been trained in the past. They learn the patterns of things in existing data - the composition of paintings, the structure of music, the senten

Humans are able to imagine things that have never existed, while machines cannot do that.

Of course, AI can bring tremendous efficiency changes in some application areas, such as writing manuals. But you still have to want to check it again to avoid any wiring errors or plumbing installation mistakes, because machines don't understand wha

So the key is that AI doesn't understand anything. It's just pattern recognition.

Therefore, if it is used for pattern recognition tasks, such as detecting cancer lesions in X-ray images in medicine, it can be very useful as an auxiliary screening tool. However, machines should not be allowed to make the final judgment.

Moreover, we should also remember that many important jobs of humans are not valued by "intelligence". For example, nurses, care workers, sanitation workers, and aircraft baggage handlers - those who load and unload luggage in the aircraft cabin. The

I prefer to be cared for by a human caregiver rather than a machine when I am old.

I was even sad to hear Matthew Cobb say that many important human jobs are not 'intellectually' valuable. At first, people said AI would help us with housework, laundry, cooking, and caring for the elderly. But in reality, what have we seen in the pa

We may have to wait for a long time to get the robot assistants that can truly complete repetitive tasks for us.

Think about the role of care workers - I myself am not young anymore. When I get old, I don't want machines to take care of me. I want human care. When I lose my self-care ability, I hope someone will help me with cleaning. I don't want it to be a ma

I ask him if it is because humans develop emotional connections.

Cobb admitted that yes, that's the wonderful place. It's about the connection between people. We don't have the same emotions for machines. I wouldn't want a "robot nurse". I just want a human nurse.

I said, even if the robot nurses are always patient, not angry, not tired, and can't make mistakes?

Cobb said yes, he understood that care work was very difficult. But because of this, we should support, train, and respect caregivers even more. Personally, I would never want to be cared by machines. I hope to be cared by professional nursing staff

"However, unfortunately, this is not yet widely achieved in Western societies. They are overworked, underpaid, and lack social status - it is not just a technical issue but a social and political one," Cobb disclosed to Sina Finance.

This article is from the WeChat public account of Sina Tech