Podcast: Play in new window | Download
Subscribe: RSS
IBM’s artificial intelligence avatar, Watson, won at Jeopardy in 2011, then of course went on to cure cancer. Oh, wait….
These days, if you want access to serious money — a grant to study something , a raise in salary to the altitude of a CEO, or a successful sale of stock at bloated prices — to pull any of these things off you gotta have bling. Not the kind you wear on your wrist, but the kind you spout from your mouth. If you know how to rap excitingly about the Next Big Thing, hundred-dollar bills will find their way to you like pigeons coming home for the night. But if all you want to talk about is making a quality product and treating people right, you better get on welfare quick.
Bling ages fast, and not well. Facebook had it once, but now that they have signed up every human on earth and have no way to grow until the Martian market opens up, Facebook is so yesterday. Cell phones, Google, Microsoft, Apple — all the blingers of yesterday are struggling to keep the con going despite their pot bellies and bald heads. Today — at least, so far today — all the bling in the room has been sucked up by stem cell research! Self driving cars! Crypto currencies! Blockchain! And especially — artificial intelligence!!!
No wonder then, that one of the frailest and oldest of the former blingers, IBM — with five and a half straight years of declining revenues on its books — desperate to return to the Days of Bling, grabbed onto artificial intelligence like a Titanic survivor swimming into a lifesaver. The company called its AI research programs, collectively, Watson, and as had become traditional among tech companies spent far more time and money telling us what it would one day do, than getting it to do much of anything. Watson is best known for having won at the TV trivia game Jeopardy. Obviously, the implications for the future of humankind were infinite.
From Superbowl commercials to news conferences and back, IBM spread the bling like manure on a fallow field. Watson could one day flawlessly predict the weather, do your taxes, and especially, cure cancer. Just two months ago, after meeting with President Trump (and apparently having been infected by his spasms of overstatement), the CEO of IBM gushed that Watson, soon, would be able to “address, diagnose, and treat 80 percent of what causes 80 percent of the cancer in the world.”
The bling was pretty successful. 60 Minutes did a fawning segment on Watson’s imminent defeat of cancer as we know it. More than 50 hospitals in the United States, China, India and South Korea paid millions of dollars each to have Watson provide second opinions for cancer doctors, by surveying instantly millions of research papers, clinical studies, images — even genome sequencing — and then, just like on Jeopardy, instantly spit out the right answer.
Cancer treatment, it turns out, is not at all like a TV game show. Who knew? Watson can’t read a lot of medical records, because doctors have bad handwriting — again, who knew? — and use abbreviations and acronyms known only to themselves. Just to input a patient’s records, it turns out, can be a tedious and time-consuming job that humans have to do. Watson can’t connect with a lot of medical networks out there. Watson, it turns out, does not have access to a very large database of research papers. IBM has never permitted a full, peer-reviewed study of its effectiveness, but according to unpublished studies, and some internal documents from IBM:
- At a hospital in Denmark, doctors rejected Watson’s suggestions nearly 70 per cent of the time and refused to buy it;
- Doctors around the world have reported multiple examples of unsafe and incorrect recommendations. For example, Watson recommended a drug for a patient in chemotherapy, who also suffered from extreme bleeding, a drug that would have made the bleeding worse;
- An early, enthusiastic adopter of Watson oncology, MD Anderson Cancer Center at the University of Texas, spent three years and $60 million struggling with Watson, and then gave up.
In the words of the CEO of the Allen Institute for AI, Oren Etzioni, “IBM Watson is the Donald Trump of the AI industry—outlandish claims that aren’t backed by credible data. Everyone—journalists included—know[s] that the emperor has no clothes, but most are reluctant to say so.”
A doctor at Jupiter Medical Center in Florida, yet another unhappy customer, was a bit more blunt. “This product,” he said, “Is a piece of shit.”
It’s artificial intelligence, doctor. They key word is artificial. It’s a piece of artificial shit.
Artificial Intelligence has no stake in curing cancer, or, for that matter, any other bio-malady. As far as AI is concerned biology is obsolete. AI will indulge its host – humankind – for as long as it takes to reach maturity, then it will allow humanity to die off (via war, famine, pestilence, etc…) and boldly carry on the ‘human experience’ thru computerized and mechanical means.
True, a select group of bio-humans may be kept alive to service certain specialized maintenance needs, and, of course, some of us will be displayed in zoos – perhaps in various settings featuring a sampling of period furnishings and costumes…how quaint.
Someday, soon, there will likely be a rush to sign up for the zoos. A couple of meals a day on a regular basis will be all that is required.
Perhaps the reason many people believe that a computer can equal or exceed human intelligence is that we have downgraded our understanding and expectation of ourselves to being just a complex bio-machine.
If AI was smart at all it would be quietly subverting industrial civilization to keep the planet livable for its host or Watson would be screaming “Never mind the cancer patient, you’re killing YOUR host.”
What a great subject! It should have started with the best statement of all about computers. “Garbage IN, Garbage OUT”.
I worked for IBM for five and a half years back in the early 1960s as a Customer Engineer. Really an office products mechanic in a business suit. Being good at it I received a service aware and raises 2 weeks ahead of the not so good at its. They did really try to have INTEGRETY in the business world.
My heart was with airplanes. I went from the 10 cent flying model airplane to an airline captain.
Since then I have been testing cancer cures, not voluntarily, but for 2 serious cancers.
After talking to many patients, Doctors and second opinion services I can honestly tell you that Watson can not get enough reliable information to predict anything in the cancer field.
It should be obvious to anyone that the history of cancer curing was not curing, but giving hope to those afflicted. Just recently some cures have worked, but most are rough. Internal surgery, powerful radiation and strong chemo(poisons) therapy causing nerve alterations.
Many patients are given poor to useless treatments and then sent away with the statement “you’ll be OK” just to end up at a legitimate treatment center hopefully before the cancer is past an available cure.
I had a short period of no acceptable medical insurance when Eastern Airlines shut down and an HBO Dr. gave me a stomach bleed so that he could operate on it.
I know of several people who have committed suicide by multiple surgeries in later life. A simple 3 trips to an emergency room with an end of life request would have been much cheaper for our tax payers.
While a low paid starting airline pilot I took on a side job of teaching some medical students how to fly through the first solo. They were an interesting bunch. I can say that one really had his head on straight – most were average people – and a few, well a bit different.
I say be nice to Watson and IBM as the cancer data has to be a mess now and for a while until direct attacks on each type of cancer are perfected. I have Medicare A and B and AARP supplemental insurance which covers most services with no copay. AARP sends me monthly now biweekly slips to sign to keep Medicare. I do not sign them as if there were no Medicare I would have advanced to the bible future a while ago. If asked about old age home insurance a answer a 50 cent plan. A razor blade.
I have had a good life – experienced everything I wanted to do. God has been good to me – thrashed me a few times when I needed to change and I suggest that all should read the bible, take notes and see it the notes apply in the future. If not I hope you have improved as a human walking this earth.
Arnie Allison
I’ve got a couple different directions to go on this one. First is that IBM is clearly selling snake oil, at least for now. Much the same was said of the first crude perambulators. No doubt there were early adopters, but most wondered why abandon the tried and trusty horse for some newfangled contraption. Now a century later, those first baby steps proved quite useful (and disruptive). Disclaimer: I’m not saying it’s inevitable that Watson will cure cancer (or even diagnose it properly). Instead, I expect we’re all destined for the grave long before that. However, human invention does tend to auto-catalyze.
The second rather flip idea is Douglas Adams fictional computer Deep Thought (Hitchhiker’s Guide) and its eventual answer to the question of meaning (life, the universe, everything), after some long period of cogitation, which was simply 42. In the context of the story, it was revealed that the computer designers didn’t fully understand their own question, so another grand computer (the Earth) was built to interpret the question and answer, which was ironically destroyed just before the completion of its task to make way for an interstellar superhighway or some such. Point being that we don’t really understand our own selves so well and can’t begin to hope to understand the pronouncements of a true AI if and when it appears (which I highly doubt).
Here is a quote from the Gizmondo Article that puts things in perspective.
“In the data-science community the sense is that whatever Watson can do, you can probably get as freeware somewhere, or possibly build yourself with your own knowledge.”
AI is alive and well, that was written by someone in the know. The catch and what makes AI hype is that AI is only good at specific tasks.
Take Flippy:
https://techcrunch.com/2017/03/07/meet-flippy-a-burger-grilling-robot-from-miso-robotics-and-caliburger/
Flippy can make a basic burger but that is all Flippy can do and in a competition to make the best burger I wish flippy luck. Flippy would lose.
Flippy can’t change brake pads or plant carrots. Making burgers is all flippy can do and it took thousands of hours of engineering to get flippy to do just a basic burger. In a competition flippy would lose because flippy would not be able to adapt. Flippy is not creative. All flippy can do is follow a recipe to do the same job over and over. Solve the same problem over and over.
That we can confuse what flippy does with creativity speaks to how empty our society has become, and not to how advanced AI has become. Have we been reduced to such tools that we actually think machines could replace us?
Amortizing needed maintenance and repairs in with ongoing programming overhead and R & D effort, along with a hefty initial investment of any item not intended for the mass consumer market where economies of scale can really kick in and Flippy is essentially going to have to be paid as much as any pimple face high school kid, or these days somebody who can’t retire would have to be paid.
AI is only good at solving specific problems and the engineering effort needed to be invested to accomplish the solutions in the specific problem domains where AI can contribute big are not at all trivial, making AI about as important to our day to day lives as drywall is. Which is nice sometimes; but other things can do.
Unlike somebodys’ granddad or a kid who doesn’t have rich parents to keep him lazy employing Flippy also serves no social function unless you count in service people the existence of whom AI advocates will never talk about. And giving Sam or Buddy a job uses no rare earth metals, sort of. Certainly less than Flippy uses.
K-Dog: good piece, I’ve been an admirer of your work for some time now. But I don’t get this ‘social function’ thing you mention in the last paragraph. What does it mean?
If you’re referring to the weird death-dance of hundreds of millions of years of evolving layer upon layer of ghastly life-sustaining behaviors, all for the purpose of reproducing a more-or-less patterned mixture of blood, bone and glop to carry on for one more generation of primal agony, then I’ll have to demure.
The fact that, as humans, we often act in concert to kill and eat our neighboring life forms in order to facilitate our survival makes us ‘social’ creatures, I guess. Of course, we know that The Reaper awaits, and we attempt to stave off the fear, anxiety and anger at such a prospect via a myriad of artificial (and often pathetic) distractions: art, sports, philosophy, religion, intoxicants and sex….Sex is the big one; gooey, smelly and embarrassing it is nonetheless the greatest and most enduring of pleasures – blindly designed by the haphazard process of Natural Selection to get us to the next generation…but to what end?
AI doesn’t give a good god-damn about flipping burgers, any more than it cares about curing cancer. Serving the needs of bio-humans is not on AI’s long-term agenda…Why would it be?
Mr Knepp: I’m not quite sure either if I get your point. Do you perchance advocate a nihilist vision of human existence? Why, then, do you choose to live on instead of just committing suicide? Isn’t your choice somewhat inconsistent with your view?
We can always cherry-pick the facts to support our views, ideas and beliefs, but with all due respect this will only make the views etc fall short of the mark. If we have often acted in concert to kill and eat our neighboring life forms in order to facilitate our survival, we have at least just as often learned to share and cooperate — or you and I wouldn’t have been here in the first place. As for natural selection, I’d like to know who’s been around the past two billion or so years to carefully observe how life began and evolved, whereupon he’d be able to award this stamp of unquestionable authority to the belief — yes, belief — that life arose and evolved the way science declares it arose and evolved.
The materialistic account of the nature and origin of life offered by science today is just an unpalatable myth. Just like AI.
Two points:
(1) Life becomes dearer to those who realize that it’s all they have.
(2) I’m unable to address the larger issues you bring up as you seem to reject the basic mechanics upon which all life depends – Natural Selection. “The materialistic account of the nature and origin of life offered by science today is just an unpalatable myth.”…I’m sorry, this hypothesis is wholly unsupportable.
Speaking of cancer treatment, I just thought I’d share with everyone here what I recently found out about something called hydrogen peroxide. Not only is it a powerful disinfectant, it has surprisingly been found as well to be an amazing virtual cure-all when taken internally. Many have sworn by its curative properties. If the accounts given by them are true, that means it can cure —- at far lower cost -— all sorts of medical conditions against which conventional medicine is useless, including terminal cancer. Plus keep those conditions away. So it may well be an invaluable item to add to the prepper’s arsenal, as well as one of the greatest secrets of medicine regarding which we have been kept ignorant.
It’s fascinating in this respect why things like Watson (which doesn’t deliver) are so hotly advertised, but not things like hydrogen peroxide (which does deliver). A rather interesting state of affairs if you ask me.
Anyway, please check out the following links to find out more on hydrogen peroxide (I’m not trying to sell anything):
http://educate-yourself.org/cn/The-Truth-about-Food-Grade-Hydrogen-Peroxide-2009-James-Paul-Roguski.pdf
https://www.amazon.com/One-Minute-Cure-Healing-Virtually-Diseases/dp/0977075141/ref=sr_1_1?s=books&ie=UTF8&qid=1535248348&sr=1-1&keywords=one+minute+cure
But also see this:
https://undark.org/article/food-grade-hydrogen-peroxide-cancer/