Original URL: http://www.theregister.co.uk/2013/12/18/ibm_predictions/
IBM predicts glorious tech future made possible by $IBMmarketing
Poor track record of scrying muddies optimistic pronouncements
Analysis IBM's latest "five in five" predictions – five things that may happen in the next five years – ignore social and economic realities to alert us to new technology that'll be prevalent (and mostly made by IBM).
The predictions were published on Tuesday, and are centered around Big Blue's major product push of "cognitive computing" (henceforth abbreviated to
$IBMmarketing). This draws upon multiple investments by the company in technologies ranging from its Jeopardy-playing Watson supercomputer, to neuromorphic chips and neural-network research.
For this year's "five in five", IBM predicts that, by 2019, online classrooms will use
$IBMmarketing to provide tailored educational programs for individual students; that
$IBMmarketing will allow storekeepers to "beat online" shops by carefully tracking the flow of goods through their business; that doctors will use DNA-specific treatments and DNA testing to provide patient care; a digital guardian that uses
$IBMmarketing will follow you round the internet to perform semi-autonomous security checks; and finally cities will start to reach out to your smartphone and/or fondleslab through the magic of
Though there is something to the idea of machine learning and/or deep learning systems being applied in broader ways, IBM's predictions seem a little long on the possibilities of
$IBMmarketing and short on the social and economic aspects that may make this difficult.
Online schools, for instance, have so far seemed to be a failure, and the shift of original darling Udacity from web learning to corporate training indicates the tech may not be viable.
As for the DNA-specific treatments, well, DNA-sequencing company 23&Me just got slapped by the US Food and Drug Administration for offering personal genome tests.
The idea that small businesses can beat online retailers is contingent on them buying in large amounts of sophisticated tech from vendors like IBM – something that many shopkeepers can likely not afford. Perhaps this prediction really means that
$MegaBricksAndMortarRetailer will beat
$OnlineRetailer, in which case we're not sure of the utility.
IBM's predictions that cities will use more advanced technology to reach out to users and inform them of things like transit delays or commercial opportunities seems like a safe bet given Google is already investing a ton of money in this area – as does the digital security agent, due to the growing problems posed by many people memorizing many passwords or using a password wallet.
These fuzzy predictions are par for the course for IBM, which has been making them for years.
If we'd believed all of its predictions from 2008 then, by now, we'd have thin-film solar cells layered on the tops of our laptops; we would be offered full DNA profiling from our doctors; we would never interact with people in stores and would instead use "digital shopping assistants"; all of our information would be reliably stored so that "forgetting will become a distant memory"; and we would be using our voice to do many of our tasks online.
Out of these predictions, the only one that has come fully true is the voice one – thanks to Apple's Siri app and Google's Now service. The other predictions were either overly ambitious (thin-film tech is still being turned into products, and DNA-profiling is possible, but the regulatory landscape is difficult), or wrong ("digital shopping assistants" do exist, but as a gimmick, and seriously pricey places will stick a comely flesh-and-blood member of the human species by your side if they think you're loaded).
And "forgetting will become a distant memory" has sorta come true, thanks to harvesters of intimate information, such as Google, Facebook and the NSA; only the most determined netizens can successfully control the logging of their lives – lest any snaps of their drunken nights out turn up during job interviews, say. There are plenty of moments that ought to be a distant memory, and for good reason.
So it seems likely that IBM's pronouncements will come true in one way or another, but over a longer time period than Big Blue imagines, and via technologies or competitors it does not yet foresee. As ever, the possibilities posed are tantalizing, but it seems the greatest deployer of
$IBMmarketing tech will be Google, rather than Big Blue. ®