Questions? (877) 643-7244
FutureNow Post
Wednesday, May. 4, 2011 at 11:31 am

Watson, Artificial Intelligence, and the Future of Websites

By Marijayne Bushey
May 4th, 2011

If you’re a regular reader here at Grok, you already know that FutureNow’s Persuasion Architecture methodology hinges on our ability to understand and predict the human behaviors surrounding decision-making, and develop business applications that allow us to apply that knowledge to your website and marketing efforts.  Many of the prospects I’ve talked to over the years thought we did this through technology, such as dynamically displayed information.  But we don’t.  A key component of how we gain insight, form hypotheses, and drive continuous improvement, is still good, old-fashioned, human critical thinking skills. Yes, we use tools like Google Analytics to collect real-time information about what people do on your site, and we use tools like OnTarget to manage our efforts and their effect on your site, but it’s not a 100% technological solution you plug in and walk away from… ultimately it’s our trained professionals analyzing the data for patterns or insight, and guiding the process, and you implementing what they recommend.

Still, these prospects and their assumptions got me thinking.  Could we ever face a 100% technological solution?  Would it be possible that someday we’d be able to have applications or programs that run behind your website, collecting and analyzing information about what people do, making predictions about what they need and dynamically displaying the required information, and then re-analyzing and refining the information based on how they react?  In theory, I knew it was possible.  I knew it was a “simple” matter of math… of developing statistical models for predicting human behavior patterns, and then programming them into your website.  Yes, when I say “simple” I jest.  In my previous life as a project manager for cancer research projects, I had seen my highly educated bosses struggling to develop models for predicting a patient’s health care choices; I knew my uncle was part of a team that developed a mathematical model for predicting hurricane behavior and landfall… and that his team had been working on the model for over 20 years; and I watched my brother – a guy who reads chemistry textbooks and says things like “that makes perfect sense” – slowly lose his sanity trying to develop mathematical models for his thesis.  Mathematical models that predict outcomes in an ever-changing system of variables are far from “simple.” Human behavior in the context of the ever-changing technology of the internet would be just that kind of system.

And then there was Watson.  Watson is the IBM computer that beat two of Jeopary’s best contestants in a February 2011 man-vs-computer showdown. I watched NOVA’s Smartest Machine on Earth prior to the actual showdown.  It revealed some of the challenges the team had to overcome in their quest to develop a machine capable of imitating human thought.  While Watson did exhibit some problems with things such as gender, puns, and other subtleties of language, he eventually managed to overcome them to win the showdown.  And from what I understand, Watson even got a job out of it: I read in the newspaper that Columbia Medical Center was going to “hire” Watson to help with patient diagnosis and treatment.  Surely, if Watson can save lives, he can help inspire your prospects to take action on your website!

But one of the most interesting things about the NOVA episode as far as I was concerned, was the divergent opinions of academics about whether or not human intelligence actually can be imitated, and more importantly, bested.  Sure, Watson looks simple on the surface, but he has one heck of a complex system behind him.  And it took years for the team at IBM to prefect him to the point where he could be a contender on Jeopardy!  Stephen Baker’s Final Jeopardy: Man vs. Machine and the Quest to Know Everything provides a history of Watson, in addition to thoughts on the subject of artificial intelligence.  While Baker recognizes that technology is heading the way of Watson, and proposes that machines such as Watson will become a part of our everyday lives, he does couch this with  “Our brains are still the most intricate, complex and brilliant thinking machines on earth. But we have to figure out how to use them in concert with the machinery we’re building.”

Pop culture plays with this idea in movies such as The Matrix, and I, Robot.  So, what do you think?  Will Watson or his baby brother one day make the likes of Bryan Eisenberg, Avinash Kaushik and Tim Ash obsolete? Or is the human capability for understanding and interpreting far too complex to be imitated?  Add your 2 cents to the discussion.

Add Your Comments

Comments (22)

  1. Maybe it’s just my ego talking here, but I don’t believe that computers will be able to best our critical thinking skills.
    Yes, we use computers to get all the analytics we use to make marketing decisions, but the data means absolutely nothing without someone thinking about what the data means and what to do with it.

    I would welcome any new computer or software that can give us more information so we can better make decisions, but please, let’s leave the thinking to us humans!

    Great post Marijayne!

    Allison

  2. Winning Jeopardy is just the beginning for this kind of tech. Someday we will be talking to these machines like they were human.

  3. @Allison Semancik – Agreed! I keep remembering what my professors told me: computers can only do what WE program them to do. ;-)

  4. We must remember something about Kubrick’ HAL9000… be attention!

  5. Hi Marijayne, but yet, people still depending a lot on the computer these days…

  6. I personally think that AI may be able to analyse data and give suggestions/predictions, but human will be the sole decision maker to take action. AI can analyse data effectively or even better than us, but unable to consider every aspects such as global trends.

  7. Disagree Marijayne. The future for computers, in my opinion, is barely comprehendable. Computers will eventually take over the tasks of human beings and will be able to “think” and display “emotions”. Maybe not in our lifetime, but definitely sometime!

  8. I think that the key is teaching computers to learn and adapt. We got where we are today through evolution and a computer needs mimic the tools of evolution in the biological world.

  9. i agree with allison. computers can never surpass humans.

  10. In my opinion a well programmed ai can compete the human brain in many ways, before all if it is a logical decision. A machine can be faster and more consequent. So simple tasks can be solved by a machine easily and also reliable but the more important ones should at least be rechecked by a human.

  11. We can have programs/scripts running behind our websites and talking decisions for us…but GOOGLE will categorize them as Malware or SPAM and know you down in their index.

  12. Predicting human behavior and especially customers behavior is the goal of many companies… The point is a program can not learn from a situation and transpose it to another situation. In a word, the algorithm can not evolve by itself efficiently. The base is maths, and we are not made of maths, we are full of contradictions and math model is just logic…

  13. @web designer – that was one of the most interesting things about Watson: that his creators eventually overcame the challenge and at lease gave him the appearance of “learning” as you describe it.

  14. Correct my if Im wrong, but is Watson considered artificially intelligent? From what I understand, he still pools together data like a normal algorithm would, but in a much more complex fashion. I think machines will pass humans one day because they can store so much more data, but they probably will have to think more non-linearly…similar to humans.

  15. Computers are able to analyze immense amounts of data however they lack the ability to take risk. A computer will analyze the options and pick the one with the lowest risk and the best outcome however without risk your business probably wont do very well. So I think that computers have one large flaw in that we have to program everything into them so therefore they are only as smart as the code we can use to describe something.

  16. don’t get me wrong, finding answers in unstructured data that way is an incredible step forward (I guess Google is watching closely this technology…) but intelligence is more than computing data, even unstructured.

  17. It frightens me to think of the rate of which our technology is progressing. It is only a matter of time until an AI is developed that will allow the processing/data retention of these machines to far surpass any human capabilities.

  18. @3D Renderings – this challenge was also noted by the team that designed Watson and overcome by the inclusion of statistical risk calculations! However, as you noted, Watson’s ability to risk was only gained by through the thought and programming abilities of the humans designing him.

  19. So many of us hate change and those same many of us want progress (including technological). Wouldn’t this partially reflect that our average ability to adapt to technological advancements is not really as good as we’d like to think? How could it be any other way than that … we will very probably NOT be able to keep up the pace. (and … this comes from an obsessive “teckie”)

  20. computers can never surpass humans but for it’s the future..

  21. I think it’s only a matter of time before we develop machines that can think critically, make intelligent decisions, and even take risk. Remember that at one point man probably thought flying was impossible.

    Companies leading the charge like Google and IBM immediately come to mind, but I think people forget about Amazon. They’re already making leaps and bounds in this area. Amazon.com probably incorporates more algorithmic and machine learning elements than any site on the web – even more so than Google.

  22. I agree with Allison’s comment. I don’t think computers have a real risk to our existence– not unless they are being operated by people who have negative intentions I mean.

Add Your Comments

 

Print this Article
Share

Mj has been proselytizing the merits of customer-centric, data-driven, continuous Conversion Rate Optimization (CRO) for FutureNow since 2007.

More articles from Marijayne Bushey

Marketing Optimization Blog
FREE Newsletter Sign-Up
send it once every: