13 Comments

Summary:

Currently much of the big data being churned out is merely exhaust. But imagine the possibilities once we figure out how to produce and process better data on the fly on a global scale. Call it Big Inference.

Screen Shot 2013-06-28 at 9.32.28 AM

These days, many of us in consumer and enterprise tech companies are working on predictive systems that provide modest but valuable augmentation of human intelligence and business processes. I think this scale of ambition is a good fit for the current state of the art in machine learning and probabilistic inference. Think personal assistants like Siri or Google Now, predictive analytics in the enterprise for churn detection and ad campaign targeting, and personalized news apps like Prismatic.

But I think that the long term story is much more exciting, and much further from our experience with synthetic intelligence to date. I believe that we are on the path to building the equivalent of global-scale nervous systems. I’m thinking Gaia’s brain: distributed but unified intelligences that gather data from sensors all over the world, and that synthesize those data streams to perceive the overall state of the planet as naturally as we perceive with our own sensory systems. This isn’t just big data–this is big inference.

The world as brain

To make this idea of a global intelligence more concrete, consider the startup Premise. As a first step toward the kind of perceptual systems that I am talking about, Premise is using various signals from the public internet as a set of massively distributed sensory organs, and then leveraging this information to develop more informative economic indexes.

Now consider what other problems such systems could solve in the coming decades. We could gain a true understanding of the climate system on a granular but global level. We could track and coordinate every vehicle on the planet, to improve energy efficiency and optimize scheduling to all but eliminate traffic jams. Or moving from vehicles to parts and materials, we could create and manage truly robust supply chains that maintain efficiency and resilience in the face of unexpected events. The possibilities go on, and are truly awesome.

Key problems in the way

To get there though we’ll have to confront a number of hurdles:

We need to gather the data. Emerging, massively distributed and networked sensors will be the equivalent of human sensory transducers like rods and cones. The rise of the Internet of Things also means that every device will be able to contribute its own data stream to a collective understanding of the current state of the world.

Much of the content of big data these days is exhaust – data originally collected for transactional or other purposes, for which mining and analysis are afterthoughts, and whose characteristics are often ill-suited to further analysis. This will certainly change, as data collection matures into a process explicitly designed to improve our peceptual and decision-making capabilities.

We need the processing power to interpret the data While it has become fashionable to note how cheap compute cycles have become, it’s certainly not the case that we can process billions or trillions of input streams in real time –especially when we need to find patterns that are distributed across many noisy and possibly contradictory sensor inputs (i.e., we can’t just process each stream in isolation). We may need to develop new processor technologies to handle these kind of astronomically parallel and heterogeneous inputs.

We need the algorithms. To actually make sense of the data and decide what actions and responses to take, we have to figure out how to extract high-level patterns and concepts from the raw inputs. There is an ongoing debate over the right approach: Most researchers will say that we need something more “brain-like” than current systems, but there are many different (and opposing) theories about which aspects of our brain’s computational architecture are actually important. My own bet is on probabilistic programming methods, which are closely aligned with an emerging body of theory that views the brain as a Bayesian inference and decision engine.

But there are other important research threads. Google is backing so-called deep learning methods, a fundamental advance in the artificial neural networks (ANNs) that promised so much in previous decades before falling short of expectations. And Jeff Hawkins’ cortical learning algorithm (CLA) claims to replicate the human brain’s ability to capture spatiotemporal patterns in arbitrary sensory inputs.

While exciting, all three of these approaches currently fall well short. More research is needed, as they say.

Scaling isn’t enough

One approach that won’t work is just scaling up the current state-of-the-art in machine learning. The brain must constantly merge its previous experience with new and diverse sensory data to quickly interpret the current situation and decide how to act. The brain doesn’t start from scratch every time it encounters a new set of observations. Instead, it leverages all of its previous inputs – in the form of a sophisticated model of “how the world works” – to quickly discover the most likely explanation(s) for new information. This is why phenomena such as priming, expectations, and framing matter so much in how we perceive our physical and social environments.

Of course, all of this coming power could be used to control and oppress just as easily as it could be used to improve the human condition. I think that world-spanning intelligence can help us to overcome some of the most fundamental challenges that we face as a civilization, but recent events demonstrate that technologies developed with one set of intentions are often put to other uses. As scientists and engineers, we need to take responsibility for our creations. This will only become more important as we create global-scale intelligent systems.

Note: This article represents the author’s own opinion and not the that of his employer, Salesforce.com.

Beau Cronin is a product manager at Salesforce. Previously he co-founded Prior Knowledge, which was acquired by Salesforce in 2012, and Navia Systems. He has a PhD in computational neuroscience from MIT.  Follow him on Twitter @beaucronin.

Have an idea for a post you’d like to contribute to GigaOm? Click here for our guidelines and contact info.

Photo courtesy Zsschreiner/Shutterstock.com.

  1. Steve Ardire Sunday, June 30, 2013

    Hello Beau,

    I generally agree with your post except the rather feeble frames on ‘Big Inference’ and ‘The world as brain’. The former is inclusive in Cognitive Computing and latter inclusive in Internet of Things ( or Internet of Everything ) like with smart sensors.

    Cognitive Computing solves complex data and information analysis problems including what you call ‘Big Inference’ and the best systems learn incrementally.

    I have no affiliation but We’re Entering the Era when Machines will “Learn” and “Think” http://www.theatlantic.com/sponsored/ibm-cognitive-computing/archive/2013/06/were-entering-the-era-when-machines-will-learn-and-think/277143/ is a decent overview and paid product pitch but IBM’s spending 10′s millions educating is good ;-)

    Finally in in addition to @IBMWatson other REAL commercial cognitive computing players ( not R&D stuff or startups like @premisecorp or @vicariousinc ) are Grok (Numenta) @groksolutions and @saffrontech

    Cheers @sardire

    Share
    1. Beau Cronin Sunday, June 30, 2013

      Steve, thanks for your comment. Something like Watson is pretty far from what I’m talking about at this point in time, though who knows how far IBM’s scientists (and marketers) will be able to push it. And the tech behind Grok has been around for some time, but that service certainly has a ways to go before it meets its own hype.

      I do think there are fundamental gaps in our current computational abilities that need to be closed before the kind of systems I’m talking about here can be built. Everyone wants to build brains, but the question is what that means. What about the brain is essential to its abilities? And what are the engineering principles that we can extract from it, and then apply to our own systems. Jeff Hawkins’s HCA (and by extension, I believe, something like Grok), at least poses a theory here.

      Share
  2. alexgulakov2 Sunday, June 30, 2013

    We are indeed evolving into a brainnet-connected superorganism, much like an ant hill Pretty good book on the subject.

    Share
  3. Doug Walton, PhD Sunday, June 30, 2013

    Well, its a nice vision and probably overall something to aspire to IMO. But, I suspect the line where you mention it could be used for good or evil is a dramatic understatement. Also, while its one thing to process all that data once its accurately entered somewhere in the global brain, the big problem is to get meaningful data in a system of that size — apples to apples comparisons and so on. Currently its a struggle to just get accurate financial and costing data entered for anything more “soft” than hard expenditures in a Fortune 500 company.

    Share
  4. Seems like the real problem is that big data is power and powerful elements are trying to get their hands on it first. Should we not come up with mechanisms to democratize access to this data? Like a wikipedia of big data?

    Share
    1. Beau Cronin Monday, July 1, 2013

      Yes – we need to push forward on social norms and policies that recognize the power and value of data, and make sure that the power to control it is in the hands of everyone with a stake in its use. I.e., all of us.

      Share
  5. Re: “We need the algorithms.” Hope it’s not the same folks that write/wrote algorithms for Wall St. “In We need the algorithms we trust…”

    Share
  6. This is really exciting. Another problem that needs to be added to the list is keeping the system alive though. A disaster that wipes our computers out could send humans back to the stone age overnight. To me, protecting the system is the #1 concern simply because of our increasing dependency on it as a species.

    Share
  7. I am not very computer savvy, but judging by the technological trends of the previous decade, I think it’s safe to say that this probably is the direction we are heading in. My interest lies in the last section of the article: how it will be used.

    In this regard, judging by the trends of the whole (or at least most) of human history, I think it’s safe to say that the monarchs and oligarchs will find a way to use it to control, oppress and exploit the masses. And if the developers of the technology don’t play along, they will be reconditioned, forced to comply or removed from the equation.

    Education is something that could help prevent this from happening, but past trends in education as well hold little hope, if any. Rational thought is usually overwhelmed by extraordinary lies. This is definitely something we need to be very careful about if we value our freedom and individuality.

    On a side note, this was an excellent read :)

    Thanks!

    Share
  8. Beau Cronin Monday, July 1, 2013

    Agreed on both counts. My fear is that this kind of approach will definitely be attempted in the surveillance and control arenas – it’s just how those folks think. But maybe we can 1) push back on that, and 2) push for development in more beneficial areas.

    As for the second point, I’ll just note that we need progress on data collection at least as much as processing.

    Share
  9. David Mason Monday, July 1, 2013

    You didn’t bring up the issue of Open Data at all. Do we really want Google and Siri to control all this knowledge, and create split brains that includes constant bias and negotiation for use?

    IMO WikiData and other databases building on the mass of Wiki freed knowledge & data are at the front of the really exciting work.

    Share
    1. Beau Cronin Monday, July 1, 2013

      This is a really important set of issues. But I wonder if the model underlying something like WikiData will extend to real time data streams? Basically, who will own the sensor networks? And what will the norms and laws be around how those data streams are used? Will they remain in private hands, or will they move into the commons?

      We have a long way to go in answering these kinds of questions, and the outcomes are hugely important to us all.

      Share
  10. Francisco Morales Tuesday, July 2, 2013

    But a big brain by itself is nothing without other members of the body to execute the brain commands….There’s where the Connected Society enters in the game as the MVP in the equation. A brain without connections is just that, a big repository of information not useful if the information cannot be communicated to others to take advantage on that information.

    The future is in both sides, big data and Connected Society.

    Share

Comments have been disabled for this post