Big Data

Big data's big disappointment: Why AI personalization is pathetic

Everyone's talking about the "personalization" of deep learning and technology like chatbots, but when are we going to benefit from it?

Image: iStockphoto/elenabs

Big data is the next big thing, but so far all it seems to do is deliver marginally better spam. And by "marginally better" I really mean "no different from the spam we got a decade ago."

Dilbert, not surprisingly, captures this phenomenon superbly well here.

Commenting on this disconnect between the potential and reality of big data, former Facebook executive (and current co-founder at Hadoop vendor Cloudera) Jeff Hammerbacher said, "The best minds of my generation are thinking about how to make people click ads. That sucks."

SEE Are you being exploited by online marketers using "tricks for clicks"? (TechRepublic)

What particularly "sucks" is that these "best minds" don't seem to be very good at it. Despite hefty investments from tech industry titans, we don't seem to be getting much closer to an AI-focused future.

You call this personal?

Indeed, anyone wringing their hands at a future of artificial intelligence replacing humans simply needs to look at the ad tech industry. For example, without fail, if I buy a book on Amazon, ads for that book follow me around the web for weeks afterward, inciting me to buy a second or third copy.

It's pathetic.

Beyond advertising things that I've already bought, a quick perusal of the "highly personalized" offers that the best minds have in store for me include: a 2015 GMC Sierra pickup truck (If there is any car on earth I hate with a visceral passion, it's a pickup truck); a free trial for Audible (to which I've been a subscriber for many years); and an Ellie Mae LOS (loan origination system) (I'm not in the market for a loan, and certainly not in the market for software to manage loans for others).

None of these remotely match who I am or my interests, despite me spending hours online every day, shoveling data into Google, Facebook, and others. I'm desperate to help these best minds get a clue as to how to personalize my online experience, but they fail...completely...over and over again.

Which may be exactly what the world wants.

Can we get more personal?

Oh, sure, we talk breathlessly about bots and the artificial intelligence revolution, but we're nowhere near the Brave New World of machines taking over.

I've had Amazon's Echo for a few months now, and appreciate being able to ask it the time or weather, but its ability to understand and respond to communication much more complex than that is limited. Unfortunately, Google didn't surpass Amazon at I/O last week with its Google Assistant and Google Home products. The use cases were pretty straightforward. As one friend described it, it was like Disney's Autopia, a future that promises much but still has us bumbling around a well-defined track while sucking on noxious fumes.

SEE How to prepare your business to benefit from AI (TechRepublic)

Will it get better? Maybe. Google, for one, is betting that machine learning will prove its secret sauce, and some, like Marco Arment, believe that an AI-infused future could supplant Apple's current device-centric dominance. In touting Google's future, however, he points out its very real limitations:

Becoming a major big-data AI services company doesn't happen completely in secret and suddenly get released to the world, completed, in a keynote. It's a massive undertaking, spanning many years, many people, and a lot of noticeable interaction with the world. (Emphasis mine)

Based on the day-to-day interaction we have with AI from Google and others, Apple may not have much need for concern. In terms of "noticeable interaction," we've been seeing Google's AI in action for years in terms of personalized advertising which, as I've said, isn't very personalized at all. (Of course, there's some evidence that tailored online advertising is ineffective even when it actually is tailored, but that's a separate issue.)

In sum, chatbots and AI may be the future, but their present realities are still woefully underdeveloped.

Also see

About Matt Asay

Matt Asay is a veteran technology columnist who has written for CNET, ReadWrite, and other tech media. Asay has also held a variety of executive roles with leading mobile and big data software companies.

Editor's Picks

Free Newsletters, In your Inbox