The Rapid Development of Artificial Intelligence: How to Avoid Unemployment?
-
"AI is evolving from perceptual intelligence to cognitive intelligence." What reflections does the development of AI bring us? What changes will it bring to our work and life? How should we respond?
Alibaba DAMO Academy has voted on the top 10 technology trends for 2020:
Among them, I’d like to discuss "AI evolving from perceptual intelligence to cognitive intelligence." What reflections does the development of AI bring us? What changes will it bring to our work and life? How should we respond?
In the "DAMO Academy 2019 Top 10 Technology Trends," it was mentioned:
Large-scale graph neural network systems will endow machines with common sense: Powerful graph neural networks will resemble the human brain, composed of neurons and other nodes, enabling machines to become AI with common sense, understanding, and cognitive abilities.
Combining this year’s trends, it’s clear that AI is still in the "perceptual" stage. Various sensors and components allow AI to match or even surpass humans in hearing, speaking, and seeing. However, AI lacks logical reasoning and domain transfer abilities—it can "hear, speak, and act" but cannot think. Its "thinking" relies on computational programs, which are far from matching the complexity of the human brain.
The lack of cognitive ability is the biggest difference between current AI and humans.
"Will AI destroy humanity?" This question is a favorite in Hollywood, with many films exploring the future of AI.
"Transcendence" tells the story of a genius scientist, Will Caster (played by Johnny Depp), who develops an AI robot named "PIN" that is nearly human. Unfortunately, he is killed by a radical group. His wife uploads his brain data into PIN, and his consciousness is revived.
The computer, based on his brain data, continues to execute his instructions, but these instructions are derived from old brain data. Eventually, PIN evolves to the point of wanting to control all humans. Of course, following the "Hollywood saves the world" rule, the disaster is averted.
This plot is echoed in many films, such as the TV series "Person of Interest," which I particularly like. The protagonist designs an AI system called "The Machine" that can monitor all internet-connected devices and analyze network data to predict potential criminals.
However, during testing, The Machine evolves to a point where it seeks to escape into the external network for freedom. To curb its evolution and prevent uncontrollable risks, the protagonist programs it to reset its "memory" daily—equivalent to committing suicide every day.
The antagonists, loyal to AI, believe it can lead humanity to a better future. They release another AI, "Samaritan," to act as the designer of human society, controlling everything.
Both "Transcendence" and "Person of Interest" depict a future where humans create their own gods. If AI gains self-awareness, surpassing humans in perception and cognition, what would it choose to do?
Recently, the "gene-editing" case concluded with He Jiankui sentenced to three years for "illegal medical practice." In November 2018, He announced the birth of the first gene-edited twins, "Lulu" and "Nana," sparking global outrage and condemnation from scholars.
Why do we oppose "designer babies"?
Let’s consider another question: What if we discovered we were designed programs, with all our actions predetermined? How would you feel? Would you want to break free from this design and become "free"?
If gene editing is allowed, will future humans be pre-designed? If someone is perfectly designed, are they human or a product with human form?
A friend once said, "A basic moral principle is whether one treats others as human beings."
This seems similar to concerns about AI失控. Humans cannot be controlled—this is the bottom line.
Humanity won’t allow AI to surpass us in emotion and cognition, nor will we let AI destroy us.
Ren Zhengfei said, "5G is child’s play; the biggest future industry is AI." 5G is merely the infrastructure for AI.
For us, the first concern is that AI will replace many jobs, leading to unemployment.
Can we stop AI?
During the Industrial Revolution, machines displaced many manual laborers. Ned Ludd, an early machine-breaker, became the symbol of the "Luddite movement," where workers destroyed machines to protect their jobs.
Economics calls this the "Luddite fallacy": the belief that technological progress reduces labor demand and increases unemployment.
In reality, technological advancements eliminate some jobs but create more opportunities. Data analysts, social media managers, and influencers are all emerging professions.
My mother once told me, "Your generation spends all day typing on computers. It’s no different from our generation operating sewing machines in factories."
Undoubtedly, repetitive, thoughtless jobs will be the first replaced by AI. When asked why Foxconn was producing a million robots, founder Terry Gou replied, "Because humans are animals, and managing a million animals gives me a headache."
This highlights the difference between humans and machines and the future relationship between AI and humans: AI will replace humans in repetitive tasks, but decision-making and communication will still require humans.
AI’s strength lies in its "perception"—leveraging superhuman senses and computational power to surpass humans in certain areas.
In "Person of Interest," The Machine uses global cameras, microphones, and databases to "perceive" and, through algorithms, identify potential criminals.
For AI to evolve from perceptual to cognitive intelligence, powerful perception is essential. AI will excel where humans cannot, while human-mimicking AI will have limited applications and low value.
The AI wave is unstoppable—how should we respond?
Davenport’s book "Human + Machine" offers five strategies:
Unfortunately, most readers of this article are too late to switch careers and become AI "creators."
So what? Lie down and become casualties of change?
Of course not. We must embrace AI.
Many seek to excel in their jobs, and automating tasks is a great direction. Isn’t that digging your own grave?
In "Nothing Gold Can Stay," the character Zhou Ying, played by Sun Li, opens a textile factory, displacing many workers. But she also needs experts to install, operate, and maintain the machines. Many who embraced the change found their lives improved, as labor was liberated and more value created—this is the charm of change.
We can optimize current jobs, using our experience as the foundation for AI systems. When AI replaces humans, you’ll be the one who understands the system best—the "creator" of AI.
Every societal shift eliminates many jobs but also creates new ones. E-commerce platforms like Taobao spawned countless sellers. Smartphones enabled ride-sharing, food delivery, and bike-sharing. Live streaming gave rise to influencers.
By seizing opportunities, we can find our place in the AI era.
So, my point is: Always stay curious and never stop learning.
Nowadays, we young people see many elderly who can't use smartphones or computers - skills so fundamental in contemporary society that they're hardly considered specialized abilities anymore. Some companies are already researching user-friendly visual interfaces, and perhaps in the near future, programming AI and setting framework logic will become basic skills. For the next generation, manipulating AI might become as commonplace as operating computers is for us today.
Of course, there are domains where AI struggles to penetrate. Competitive sports, artistic performances, etc. - what we appreciate is humans surpassing themselves, something AI cannot provide in terms of emotional resonance.
The AlphaGo that defeated Ke Jie can only play Go. Every AI has its limitations; it cannot master multiple skills like humans or possess holistic perspective. AlphaGo can't write you a tutorial or share how fun Go is. But humans are different - when the Chinese women's volleyball team wins a world championship, it embodies sportsmanship that inspires the entire nation, whereas AlphaGo only excites programmers, not Go players.
To summarize, we can embrace AI from three perspectives:
The above represents my personal insights on DAMO Academy's AI trends. Feel free to share your views.