AI – Artificial Intelligence – is grabbing news headlines right and left. But what is it, exactly?
Webopedia tells us, simply:
“Artificial intelligence is the branch of computer science concerned with making computers behave like humans. The term was coined in 1956 by John McCarthy at the Massachusetts Institute of Technology.”
Encyclopedia Brittanica says:
“Artificial intelligence (AI) [is] the ability of a digital computer or computer-controlled robot to perform tasks commonly associated with intelligent beings. The term is frequently applied to the project of developing systems endowed with the intellectual processes characteristic of humans, such as the ability to reason, discover meaning, generalize, or learn from past experience.”
“Artificial intelligence (AI, also machine intelligence, MI) is intelligence displayed by machines, in contrast with the natural intelligence (NI) displayed by humans and other animals.”
Not every AI “looks like” a robot. Your smartphone has apps, like Navigation, that can learn from experience.
The real AI is the computer programming loaded into it. IEEE Spectrum reported, on February 1, 2017:
“Last week, we reported on an algorithm that recognizes skin cancer as well as the world’s best dermatologists. That computer program was trained using 130,000 images from more than 2,000 diseases. It, like most artificial intelligence (AI) breakthroughs, relied on big data.”
Like humans, AIs are capable of logic, math, semantics (meaning), making generalizations, and changing behavior based on previous experience.
Unlike humans, AIs have no ability whatsoever to feel emotions, dream or employ intuition. A robot never operated “on a hunch.”
Speaking of operating, IEEE Spectrum is keeping score on how AIs stack up, medically, to NIs (human doctors).
Out of three categories – “A.I. Wins,” “It’s a Tie,” and “Doctors Win,” the AI comes out on top. You might even call it a sweep.
Not surprisingly, human doctors surpass their mechanical counterparts at General Diagnosis of common diseases. A Harvard Medical School study found a 72% success rate for your family doctor compared to 34% for the computer.
On the other hand, in China, an AI performed as well as three human ophthalmologists in spotting childhood cataracts, out of 50 cases. Both types of “brains” were equally effective in diagnosing brain and skin cancers.
Chinese Robot is First AI to Pass Medical Exam – Photo by China Daily
As for surgery, AIs can make sewing-machine sutures – but do you really want a machine, without human supervision, cutting into you?
AI skill begins to surge ahead of human doctors where any complex computation is involved. For example, a machine learning program was 90% accurate – almost perfect! – in predicting which patients with mild cognitive impairment would develop Alzheimer’s disease. AIs also have very high success rates in predicting pneumonia, heart attacks, strokes, and infant autism.
For a new and advancing technology, AI seems to be coming right along.
What’s the Downside?
Healthcare providers claim that AI is already lowering the cost of their services, by automating tasks now done by humans. This means that people doing those jobs will need to develop another occupation.
In a September 29, 2016 article by Spencer Nam titled, “Will AI really lower the cost of healthcare? A lesson from a hospital without beds,” the author explains:
“…why computers will not replace doctors in our lifetime. The current patient care processes and the liability model cannot easily be automated and taken over by a machine, and physicians will remain integral to healthcare..A group of oncologists in New York City are researching the potential of IBM’s supercomputer, IBM Watson. Rather than look to Watson as a replacement, they are researching its potential to enhance the work that they do, by assisting oncologists with research and treatment recommendations.”
Despite all the acclaim AI is getting, hardly anyone focuses on the actual expense – the cost – of developing this ultra-high technology. Public attention is drawn instead to the benefits of AI.
Nam continues by saying, “the cost of involving a supercomputer in patient care has not been deeply discussed. Some people may logically conclude that just as personal computers have become affordable to a broad population, supercomputers that provide healthcare services will also become affordable over time. Unfortunately, this is not the case.”
This author agrees that “supercomputers should be providing solutions for nurses and retail clinics.”
It is the front line troops, after all, that always need support, and the generals who claim the glory.
While “keeping score” between AIs and NIs is All-American, perhaps patients will be better served by cultivating the attitude of partnership between the two forms of intelligence.