Artificial Intelligence Meets Human Intelligence
Written by Paul Siluch
February 9th, 2023
AI – or Artificial Intelligence – is taking the world by storm. This is thanks to the rapid emergence of ChatGPT, a natural language program that answers questions in plain English. And in great detail. This free program has been tried by over 100 million users in just two months, shattering the old record of nine months to reach this number held by TikTok.
It is the fastest growing consumer application in history.
People are very intrigued, to say the least.
The idea of machines and robots with superior intelligence dates back as far as the ancient Greeks. In 700 BC, the myth of Talos, a bronze robot who protected the island of Crete from invaders.
(image from artist Ray Harryhausen).
The Greeks also invented the world’s first calculator called the Antikythera mechanism around 100 BC to predict eclipses and the motion of the known planets.
It was the world’s first computer.
The Antikythera Mechanism Research Project
We have been moving towards thinking machines for a very long time. We are also rather afraid of them.
The motion pictures of the 1920s, specifically Metropolis and Charlie Chaplin’s Modern Times envisioned futures where machines ruled.
The 1968 movie 2001: A Space Odyssey featured a sentient computer named HAL who turned on his human masters with the famous line “I'm sorry, Dave. I'm afraid I can't do that.”
And of the course, there was Skynet from the 1984 Terminator movie. Skynet was an artificial intelligence program built to run America’s defense networks, only to destroy humanity when it decided we were its biggest threat.
A Brief History
The modern quest for artificial intelligence – AI – is most associated with Alan Turing, the mathematician who cracked the German Enigma code-breaking machine in WW2. He proposed the Turing Test which would determine if a machine’s ability to exhibit human behaviour was indistinguishable from a human’s.
In 1996, IBM unveiled Deep Blue, a chess program that beat the world champion Garry Kasparov in 1997.
Human chess ability may have peaked in 1997, but AI was just getting started. A program called Stockfish was the reigning AI chess champion in 2017 until Google created its own AI system called DeepMind. DeepMind wrote its own chess program called AlphaZero which beat Stockfish after just 24 hours of training. Another version of AlphaZero called AlphaGo beat the world champion of Japanese Go, once considered as a game that could never be taught to a computer.
Defining Artificial Intelligence
Intelligence is defined as “the ability to acquire and apply knowledge.”
Scientists have been trying to recreate human intelligence for decades and, while we appear to be getting closer, we are still a long way from machines that think like us. This is because human intelligence can create entirely new knowledge while artificial intelligence merely organizes things we already know. Games like chess that rely on rules and systems lend themselves to machine learning, which tests these rules at incredible speeds to reinforce what works and what doesn’t.
If AI can’t think in the classical sense, it can reorder data that we may not see when we are evaluating a problem. And it does this much faster than the human brain can. It may offer a new perspective or even a new solution we don’t see. It could help us decide which new car to buy because it can aggregate far more information than a human mind can.
AI may not think, but it sure can look like it does.
ChatGPT takes AI another step forward. ChatGPT answers questions in a far more conversational format than the current search engine methods. Here’s an example.
If I type in “What is love?” to my Google search engine, I get
- A video for the song “What is love?” by the artist Haddaway (never heard of him)
- A definition of love from a psychology website
- A Wikipedia entry on all the songs ever written called “What is love?”
I now type in “What is love” to ChatGPT and I get an answer in plain English without links to websites or ads:
“Love is a complex and multi-faceted emotion that can refer to feelings of affection, attachment, and deep caring for another person. It can also refer to actions that demonstrate those feelings, such as acts of kindness or sacrifice. Love can take many forms and can be experienced in many different ways, but at its core, it is generally considered to be a positive and uplifting emotion that brings people closer together.”
In 1942, Isaac Asimov published his Three Laws of Robotics to govern robotic minds. Written as science fiction, they have been closely considered by every AI scientist since:
A robot may not injure a human being or, through inaction, allow a human being to come to harm.
A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.
A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.
Inventors like Elon Musk worry a great deal about AI, even as they push forward with it. OpenAI - the creator of ChatGPT – has in its charter safety protocols to be followed in the pursuit of AI: safe exploration, avoiding negative side effects, and shutting down in the event of a major conflict.
Google’s DeepMind tests to make sure its programs cannot disable their own kill switches.
Even though Alphabet (NASDAQ GOOGL) had the early start with its DeepMind program, Microsoft (NASDAQ MSFT) has jumped into the lead.
Microsoft was an early investor in OpenAI, the open organization working on AI systems, and has since invested over $10 billion to control the result of it – the program ChatGPT. Microsoft is now integrating ChatGPT into its Bing search engine and its Teams program.
Google Search is the world’s dominant search engine with over 92% market share (source: Oberlo), so it would have a lot to lose if a better search method came along. Microsoft’s Bing has a paltry 3.4% market share of global search and so has everything to gain by adding artificial intelligence and human-style answers to its search queries. It did exactly that this week.
Google unveiled its own AI interface called Bard days later, only to watch in horror as Bard answered its first query wrong. The stock promptly fell $8 per share while Microsoft’s shares rose. Baidu (the Chinese search engine) and Salesforce (NYSE CRM) have announced ChatGPT-like programs, while other companies are already working deeply on AI. Amazon (NASDAQ AMZN) uses AI in its Alexa device, IBM (NYSE IBM) has Watson, and Apple (NASDAQ AAPL) uses AI extensively in its camera and other parts of the iPhone.
- If Microsoft can gain an early lead and expand the use of Bing in global searches, it stands to gain. The shares were upgraded by two research firms this week because the company said if it can add just 1% of market share, it would be worth an extra $2 billion in ad revenue per year.
- Alphabet has the most market share to lose, since it is the dominant player. However, you can never count Google out. DeepMind has enabled dramatic gains in understanding how proteins fold – a boon for the life sciences – and Google Translate (another Google AI offshoot) translates over 100 billion words per day in 109 languages. The company will be very competitive in AI.
- It is early days. We won’t see much revenue from plain language interfaces and other AI programs for a while, but we will see a great deal spent developing them. These programs demand a huge amount of processing power in the cloud, so we will be watching chipmakers Nvidia (NASDAQ NVDA) and Advanced Micro Devices (NYSE AMD), as well as cloud network management company Arista (NASDAQ ANET) closely. It may be like the California gold rush, where the sellers of picks, shovels, and jeans made the most money. Not the miners!
- This will have a huge impact on how we learn as humans – in school, at work – because we are suddenly able to get almost any question answered quickly and in a way we can understand. Will there be job losses? There always are with every technology shift. But there will also be many new jobs created that we cannot even fathom yet.
- My partner Lisa Hill says she heard the introduction of AI compared to the invention of refrigeration. The ability to lower temperature was a huge boon to society, but it did not take off until companies like Coca-Cola (NYSE KO) and Safeway figured out how to use refrigeration to its fullest potential.
- Lisa also points out that we are already seeing stock promoters listing “can’t miss” AI companies. Like other fads that came and went - fibre optics, dotcom domains (remember buy.com?), ‘web 2.0’, graphene, cobalt, and more recently lithium – expect to read about new AI companies that are on the verge of “replacing Google!”