Is machine learning the next commodity?


It’s not every day you can witness an entire class of software making the transition from specialized, expensive-to-develop code to a general-purpose technology. But that’s exactly what’s happening with machine learning.Chances are, you’re already hip-deep in machine-learning applications. It’s how Google Photo organizes those pictures from your vacation in Spain. It’s how Facebook suggests tags for the pictures you took at last week’s soccer match. It’s how the cars of nearly every major automaker can help you avoid unsafe lane changes.

And it’s also the start of something even bigger.

Machine learning – which enables a computer to learn without new programming – is exploding in its ability to handle highly complex tasks. It can make houses and buildings not just smart, but actively intelligent. It can take e-commerce from a one-size-fits-all experience to something personalized. It might even find your next date.

Driving this surge of machine-learning development is a wave of data generated by mobile phones, sensors, and video cameras. It’s a wave whose scope, scale, and projected growth are unprecedented.

Every minute of every day, YouTube gains 300 hours of video, Apple users download 51,000 apps, and 347,222 100,000 Tweets make their way into the world. Those stats come from the good folks at Domo, who call the time we’re living in “an era where data never sleeps.”

Intel Capital's Sanjit Dang

Intel Capital’s Sanjit Dang

Until now, the hot topic of conversation has been how to analyze information and take action based on the results. But the volume of data has become so great, and its trajectory so steep, that we need to automate many of those actions. Now.

As a result, we expect machine learning will become the next great commodity. In the short term, we expect the cost of advanced algorithms to plummet – especially given multiple open-source initiatives – and to spur new areas of specialization. Longer term, we expect these kinds of algorithms to make their way into standard microprocessors.

Marc Andreessen once said software is eating the world. In the case of machine learning, it will have a very large appetite.

Proprietary becomes open

To understand the potential of machine learning as a commodity, Linux is a good place to start. Released as a free, open-source operating system in 1991, it now powers nearly all the world’s supercomputers, most of the servers behind the Internet, and the majority of financial trades worldwide – not to mention tens of millions of Android mobile phones and consumer devices.

Like Linux, machine learning is well down the open-source path. In the last few months, Baidu, Facebook, and Google have released sets of open-source machine-learning algorithms. Another group of high-tech heavyweights, including Sam Altman, Elon Musk, and Peter Thiel, have launched the OpenAI initiative. And universities and tech communities are adding new tools to the mix.

In the short-to-medium term, we see three outcomes from this activity. First, companies that need to integrate machine learning into their products will do so inexpensively – either through their engineering teams or third-party vendors.

Second, a three-tier system of available algorithms will establish itself. At the bottom layer will be open-source code. In the middle will be code with greater capabilities, available under license from Amazon, Google, Microsoft, or one of the other big players. At the top will be the highly prized code that keeps these companies competitive; it will stay closely guarded until they feel it’s time to make it available widely.

Finally, we forecast a flurry of merger, acquisition, and licensing agreements as algorithm providers look to grow and defend their positions. We also expect more specialization as they attempt to lock down various markets.

In fact, that process already is well under way.

Smarter buildings & commerce

For all the talk about smarter homes and buildings, today’s technologies aren’t nearly as intelligent as they could be. Yes, they can collect data and operate within confined parameters. But they can’t adapt to the way you live your life.

If you get a new dog, for example, fixed-intelligence devices can’t tell the difference between the two of you. If your calendar shows you working from home, these devices won’t think to disable your security system without asking.

Fortunately, that’s changing. Startups such as Nuro Technologies, for example, are pairing sophisticated sensors and self-learning networks for in-home applications. Think of the sensors as mini iPhones in and around your house. You can download software into them – fire sensing, irrigation control, security and more – the same way you load apps into a phone.

Commerce is also a big opportunity for machine learning. Maybe the biggest. One of our portfolio companies, Vizury, uses machine learning to help companies display only the online ads you want to see. Awarestack is another great example: it uses data about how and where you park a car to create algorithms that can help you get around more efficiently.

Then there’s Dil Mil, an online dating app very popular in the South Asian community and growing rapidly. Unlike conventional apps that use the data they collect to make a romantic match, it looks at social behaviors – such as posting on Instagram, Facebook, and Twitter – to find the best possible match. All in real time.

Next stop: silicon

If the Linux of the 1990s illustrates the long-term impact of machine learning, the laptop and desktop machines of the 1980s point to their final destination. In a word: silicon.

Just as modems and graphics cards made their way into microprocessors and motherboards, so will machine learning software. There is simply too much data through which companies need to sift, too many actions they’ll need to take, and too many good algorithms already available.

It’s going to be an exciting time.

A director at Intel Capital, Sanjit Dang drives investments in user computing across the consumer and enterprise sectors. He has also driven several investments related to big data, the Internet of Things, and cloud computing.

Gadgets Like Fitbit Are Remaking How Doctors Treat You.


Dr. Eric Topol, a cardiologist at the Scripps Clinic in San Diego, knows when his patients’ hearts are racing or their blood pressure is on the rise, even if they’re sitting at home.

With high-risk patients hooked up to “personal data trackers” — a portable electrocardiogram built into a smartphone case, for instance — he and his researchers can track the ups and downs of patients’ conditions as they go about their lives. “It’s the real deal of what’s going on in their world from a medical standpoint,” says Topol, whose work is part of a clinical trial. “The integration of that with the classical medical record is vital.”

Similar efforts are underway around the country, as physicians and other providers seek to monitor patients remotely through new technologies, aiming to identify problems early and cut costs and inefficiencies in the healthcare system. The approach is a key focus of the nation’s Affordable Care Act, and the influx of data from internet-connected devices could be a valuable tool for health systems, helping them to maximize resources and target interventions toward patients who will benefit most. It’s also a huge potential boon for companies that manufacture these technologies and have the know-how to store and wring value from the data they generate.

Similar efforts are underway around the country, as physicians and other providers seek to monitor patients remotely through new technologies, aiming to identify problems early and cut costs and inefficiencies in the healthcare system.

Already, mobile apps, scales, and activity trackers that beam data they collect to the cloud are helping some doctors and hospitals keep tabs on their patients and inform treatments. Insurance and electronic medical records companies are investing in and partnering with tech outfits like RedBrick Health and Audax Health, which encourage consumers to use activity and health tracking tools and upload the data to their platforms.

Apple, Adidas, Samsung, GPS maker Garmin, audio tech company Jawbone, and gaming hardware manufacturer Razer are developing products that measure biological functions at ever faster clips. Startups across the country are creating gadgets such as pill boxes that can monitor whether patients are taking their meds and under-the-mattress sensors that measure heart rate, breathing and movement. Microsoft HealthVault — Microsoft’s web-based electronic health records platform — lets doctors access data from fitness trackers like Fitbit or Nike+ Fuel Band and glucose and heart monitors that patients have uploaded themselves. It’s an attempt to create a one-stop shop for health information.

Many medical professionals have been slow to embrace the concept of patient-generated data — partly because many are skeptical of information they don’t collect themselves and because many consumer-grade apps and gadgets aren’t approved by the U.S. Food and Drug Administration, the agency that regulates medical devices. In addition, some doctors and other patient advocates are concerned that internet-based systems aren’t secure and that patient privacy might be breached, intentionally or not. But there are signs that resistance to patient-generated data systems is eroding as the healthcare system shifts to focusing on outcomes, and institutions look to web-based solutions to expand their reach and save money.

Thinking Outside the Silo

Last week, Practice Fusion — the fourth largest vendor of electronic medical records in the country, according to Bloomberg Businessweek — announced a partnership with AliveCor, Inc., maker of a smartphone heart monitor, and Diasend, an online diabetes management system. When patients approve sharing data from these FDA-approved services, their information will start flowing into their Practice Fusion medical records. The company plans to integrate more devices that help consumers track their health, according to Matt Douglass, the company’s co-founder and vice president of platform.

Scripps’ Topol called the announcement an important but “baby” step toward making data-powered medicine a reality. “It’s the future,” he said. “But we’ve got a long way for this to become routine.” Integrating data into medical records can be clunky. Topol’s patients, after all, must still email him screenshots of their information before it can be put into their records.

Companies like AliveCor and Diasend require FDA clearance for medical use because they provide diagnostic services. But others — like Nike+ FuelBand and Fitbit, which work essentially like pedometers, or Wellframe, an app that guides patients through a cardiac rehabilitation program — are meant to foster healthful habits. For now, that distinction saves companies from the drawn-out and expensive process of applying for FDA approval.

Integrating data into medical records can be clunky. Topol’s patients, after all, must still email him screenshots of their information before it can be put into their records.

“Right now, there’s a void in the industry in terms of what do you do with this information,” says Tapan Mehta, the chief of global healthcare marketing for networking giant Cisco. “How do you take this data and synthesize it and make it into knowledge, which can then be used at the point of care?”

Another roadblock to making all this patient-generated information medically relevant is that it’s in silos controlled by the companies that collect it. Plus, analyzing it can be pricey. What’s needed, some experts say, is a system that aggregates and distills data into easily digestible nuggets of information for both patients and their doctors. For consumers to buy in, the interface needs to be as simple as signing into services with your Facebook account, says Guido Jouret, Cisco’s Internet of Things general manager.

Google Health was an early attempt at integration that failed because uploading the data was a hassle, he says. Now Practice Fusion is making a go of it. The company already brands itself as a “physician-patient community,” allowing patients to directly manage their health and find providers. Integrating consumer-grade health products was a logical next step. For now, patients must come into their doctors offices to upload data to the platform wirelessly through the cloud, but there are plans to let patients upload their own data from home in the future. The idea is to leverage the power of the internet to increase social interactions and productivity and provide users seamless, on demand data access from any device.

“If you look to 2020, there’s no way electronic medical records are not running primarily in the cloud,” Douglass recalls Ryan Howard, his co-founder, saying when he approached him in 2005 and sold him on the idea of starting a web-based electronic health records company. “All medical information had to be instantly accessible.”

The Privacy Problem

In the long-term, Douglass says, the company could develop “fairly complex algorithms that are looking at trends across patient populations — who’s healthy or who can be healthier and whether recommendations are actually making them better.” Like Google and Facebook, the San Francisco-based startup acts as a marketplace for information. Its services are free to the more than 100,000 medical professionals who use its product. The company makes money by partnering with diagnostic labs, imaging centers and drug companies and through targeted advertising.

Photo: Jon Snyder/WIRED

Practice Fusion has to play by federal rules governing patient privacy under the Health Insurance Portability and Accountability Act. It says all its data is aggregated and stripped of anything that would identify patients. Privacy advocates are concerned, however, the federal privacy law doesn’t apply to the growing volume of data produced by many health consumer apps and devices. Even outside a medical context, web titans like Google, Microsoft, Amazon and Facebook have faced criticism when using their customers’ activities to target them with ads for products and services. Using health data to target patients makes the stakes even higher, some privacy advocates argue.

Privacy advocates are concerned, however, the federal privacy law doesn’t apply to the growing volume of data produced by many health consumer apps and devices.

“The big concern with services that collect or aggregate health data from multiple sources is that many of them will not be covered by health privacy laws,” says Deven McGraw, the director of the Center for Democracy and Technology’s Health Privacy Project. “Consequently, how they collect and use health data is going to be governed by the companies’ internal privacy policies, which they write.”

Bob Kocher, a partner at venture capital firm Venrock and a former special assistant to the President for healthcare on the National Economic Council, says medical data today is more secure than ever. “We lost paper charts all the time,” he says. “Now we actually know which servers they’re on, and we can even document if there was a breach.”

Plus, he says, health data hasn’t yet proven all that valuable for hucksters. The bad guys don’t care about your health: They want your identity, and they can piece that together from your birthday, social security, e-mail and address — information they can get from variety of sources including bank statements, he says.

For physicians, on the other hand, the information can be invaluable. “We’re getting data that we’ve never had before,” Topol says. “It’s quite extraordinary.”

Bluetooth group ushers in updated Bluetooth 4.1.


The Bluetooth Special Interest Group (SIG), the regulatory body responsible for the standard, announced on Wednesday its release of an updated version of the specification, Bluetooth 4.1. This is the first new update to the standard in nearly four years. Bluetooth has become a familiar and fundamental word in the vocabulary of device interconnectedness and “Internet of Things,” as the technology standard that enables information exchange between wireless devices. Announced by the Bluetooth Special Interest Group, Bluetooth 4.1 brings improvements, enablements, and developer support benefits. Also on Wednesday, Suke Jawanda, Bluetooth SIG chief marketing officer , blogged “Improving Usability extends the brand promise to consumers with an ‘it just works’ experience. This spec is engineered with several new features to make it work seamlessly with popular cell technologies like LTE, maintain connections with less frequent manual reconnection, and deliver a more efficient data exchange.”

For device users, Bluetooth 4.1 will show improvements in the form of easier connections. Devices can reconnect automatically when in proximity of one another. The user leaves the room and come back to find the two devices that were recently used reconnected without any intervention.

Device users can also expect improved data transfer. Data-gathering sensors in devices while on a bike ride, run, or swim, will transfer that data more efficiently when the consumer returns home.

For developers, Bluetooth 4.1 will support Bluetooth Smart products and solutions with “dual-mode topology” and “link-layer topology” software features. What that means is that application developers as well as product developers can think about creating products that take on multiple roles. With 4.1, one can think about behavior as a Bluetooth Smart peripheral and also as a Bluetooth Smart hub. A smart watch can behave as a data-gathering information from a heart rate monitor, but at the same time behave as a peripheral to a smartphone, showing notifications from the phone. According to SIG, “As the Bluetooth Smart ecosystem grows, the Bluetooth SIG expects more solutions to play both a hub and peripheral role. Bluetooth 4.1 delivers this type of flexibility to Bluetooth Smart devices and application developers.”

The group regards the new update as “an important evolutionary update to the wireless standard.” The last update in 2010 was instead considered as a revolutionary update in the introduction of Bluetooth Smart (Low Energy) technology. “Bluetooth Smart technology put us on a rocket ship of growth, with Bluetooth annual product shipment projections skyrocketing to more than 4.5 billion in the next five years,” said Jawanda.

To be sure, the standard for wireless interconnections has become a major presence in devices and services used every day. The Bluetooth SIG, a trade association, now counts over 20,000 member companies and oversees the development of Bluetooth specifications, and promotion and protection of the Bluetooth brand.