I thought I’d share this post that ManyWorlds CEO Steve Flinn shared recently after we had a chat with Michael Schrage – who is always such a pleasure to be around…
By Steve Flinn
As we ushered in the New Year, we were fortunate to have recently had the opportunity to spend some time with Michael Schrage, perhaps the world’s foremost authority on business innovation, and, of course, a ManyWorlds Thought Leader. Our discussions with Michael were very wide-ranging as usual, but we eventually got on the topic of what we thought were the technical areas that will have the most impact on businesses over the next few years, and yet are currently very under-recognized. Here are three big ones that we kicked around:
1. Massive parallelism in information acquisition.
The reduction of scarcity in the computing world has already transformed myriad processes – we no longer worry about conserving scarce memory, storage, bandwidth, etc. We tend to forget that just a decade or two ago those were big bottlenecks. Cycle times are dramatically reduced when the marginal cost of formerly scarce resources plummet. In other words, waste makes haste! Now this same philosophy is being applied to experimentation. Massive parallelism of experiments, or more broadly, information gathering, has already had a profound impact on the biosciences, and now is poised to revolutionize materials science in general. While nano-tech gets all the buzz, ‘waste makes haste’ is the revolution that will hit first.
2. Inferencing from massive information.
The amount of information that is being generated, whether from ‘waste makes haste’ above, or from commercial transactions, market data, etc. is growing exponentially. Making inferences from this data remains a challenge. Coming to the rescue is a set of new inferencing techniques, most notably statistical learning models, that have just been mathematically established in the past five years or so. Some of these models are based on the biggest advance in the predictive modelling arena in 500 years! The Greeks gave us deduction, the Renaissance brought us induction, and now statistical learning theory brings us transduction. These new techniques promise insights from data that were previously inaccessible.
3. Adaptive systems and processes.
We all want systems and processes that adapt to our preferences and interests over time with use. By tracking our system and process usage behaviors, along with those of others to whom we have an affinity, systems can learn to become more and more effective. This approach was tried a number of years ago under the umbrella of ‘intelligent agents’, but the data gathered was too sparse, and the means of behavioral information capture too obtrusive to be effective. Now those barriers are rapidly falling, and application of statistical learning approaches described in ‘Big Idea 2’ above, will ensure adaptive systems and processes will become the norm.
Each of these areas is significant in its own right, but you can also tell from my brief musings that they will tend to reinforce one another – amplifying their impact. Transformational, yet non-obvious, technologies such as these three big ideas are the ones that end up driving the transformation of industries, and making or breaking businesses. As always, drop me a line if you would like to discuss.