Thursday, May 15, 2008

Bruce McCabe on Future Technology

Every couple of years I have the pleasure to hear a talk by Bruce McCabe, and each time he presents a riveting vision of our technological future.

McCabe is a Sydney-based industry analyst who, through his company S2 Intelligence, gives vision to a number of large organisations, including all Australian governments at State and Federal level.

The following increasingly terse narrative is drawn from my notes of his speech, in which he canvassed a number of technological developments that are already present, and will be 'disruptive' [to businesses] in the near to medium term. My comments (or elaborations) are in square brackets. Throughout, I noted that most of these developments have very significant privacy implications. Any errors or omissions, blame me.

The sections below are Video, Storage, Voice, Human Networking, Image Processing, Spacial Media, and Sustainability Monitoring. I leave it to you to surf the sites mentioned; I have not yet had time to go through them all.

Although currently mainly consumer-driven, video over the internet will be increasingly geared to business needs; it will account for 98% of internet traffic in a year or two. Technology is already available that can index the content of video clips, and so that content of videos will become searchable. Use of video will become more structured [and commodified] to the point where they can be treated similarly to text-based objects, including cutting, pasting, and hyperlinking to the middle of a clip. Moreover, there will be automated [computer-, not human-directed] analysis of clips - news in particular. Reuters is starting video news feeds that are directed specifically to machine analysis. Bulletins will be mined for meaning; for example, a report on a given company could be analysed for sentiment, which could feed into automated share traders. (I find this concept particularly insidious, as any automated trading can exaggerate the volatility of share markets, and this mechanism has potential to disrupt markets on flimsy bases.)
Walmart is currently working on a system to automatically all shopper movements in its stores, for use in marketing analysis.
Some (lead) police departments will have all officers video recording their full day by 2010.

Portable devices (phones in particular) will take up Terabyte storage, to the point where by 2025, all movies ever made (including Bollywood) could be stored on an iPod-like device. There will be a very steep curve in the takeup of storage over the next 5-10 years, to the point where people will stop deleting anything: they _can_ keep and index everything, and the [labour] cost of deletion will be too high to be worthwhile.
[Yet according to US-based industry analysts Forrester (see here), the cost of storage equipment and management software currently consumes 10% of IT budgets, and will increase 4% this year. So I would say storage maintenance will remain a significant issue, deletions or no.]

Stress analysers are working their way into call centres. Already two UK insurers are using voice analysis specifically as a component of their risk assessment. Bruce said that Kishkish, a company providing add-ons for Skype, already has a consumer-quality "lie detector" available [which I would suggest is of limited merit]. The US army has just started handing out portable voice analysers that can operate with a variety of Iraqi languages, with an 80% (?) success rate (after baselining each subject with 20 neutral questions).
Links: Kishkish

Human Networking
As LinkedIn is the most successful business networking tool, so other tools will be developed that will automate networking processes, to the point where a social network map could be built simply by analysis of the contents of email boxes. Bruce suggested such tools could be a boon for marketing in areas such as recruitment.
But that's only the beginning. Bruce depicts a point (in a process which has already started) where machines will automatically mine the web for all data about a given person.
There's more. combines machine mining with a wiki - to enable people to add their own comments to a store of information about a particular person. There was a suggestion this will greatly fuel reputation management as an industry.
Links:;;,, LinkedIn

Image Processing
Image recognition married with social processing. There could be great value in marrying computerised image recognition with social processessing - say, having a few people validate an image [or identity-related information]. A California university was mentioned which claims 95% certainty on _partial_ face recognition.
Image recognition is such that by 2009, a service could be provided that tags with location any photo that includes landmarks as significant as a building [? - methinks this is optimistic].

Spacial Media

Disruption is in the chips. At $1 per GPS chip, GPS (and RFID) to become ubiquitious. One billion GPS users by 2016. Further, with expansion totools such as Google Earth, there will be 3D views to every street, to the point where everything now achievable in Second Life could be done in a Google Streetview type environment. This has great applications not only academic (eg museums), but also commercial and intelligence-related. By 2018, asset audits to become obsolete.
Links: Second Life, Google Earth

Sustainability Monitoring
Project Vulcan: carbon emission monitoring, spacially navigated, to the discrete level of buildings, daily updated. Carbon labelling at an asset/product level.
Links: Project Vulcan

Now, much of this I would take as nigh-on achievable with current technology, but so much would rely on the degree of uptake. I discussed this briefly with Bruce afterwards, and he acknowledged that a lot of what he talked about was likely to be taken up in a leadership context, ie by relatively few, key organisations. My point to him was that we never could have foreseen the mobile phone phenomenon: a) how ubiquitous they would become in a relatively short time; and b) how such a mass consumer uptake would fuel a host of technology spurts that might not otherwise have happened without such a gross device commodification.

Bruce also expressed to me a high degree of optimism that the takeup of sustainability and carbon-monitoring technology would be pretty much led by consumer (/voter) demand. My feeling, however, remains that although by now people are dead keen to see something done about global warming, a) they are unlikely to take too much action themselves; b) they (as a mass) may well baulk when gross personal costs or lifestyle changes are at stake.

My imagination was most definitely fired by Bruce's prognostications - as it has been each time I've heard him. But although he can give good outlines of what could be done with technology (with little to no leap from today's capabilities), how it actually pans out is, I feel, still up for grabs.

No comments: