Dawn of the Emotional Machine

sexbots I was at a conference a couple of weeks ago, and a group discussion broke out about machine intelligence. It was thought-provoking, and it’s lingered. Today I saw this article posted by @lisapadilla about sex robots replacing men in relationships. Certainly it is a topic that will elicit a full spectrum of jokes and guffaws, but there is a sober angle to this that I find interesting. As Lisa asked, “can you emotionally bond with a robot?”.

The short answer, I believe, is yes.

To explain why I believe this is possible, I want to detour and look at something unrelated, the human brain. I read this great article about the brain boom that human evolution experienced. In summary, humans split off from apes and chimps as much as 8 million years ago but 3 million years ago our brains went through a dramatic expansion.

Three million years ago the human brain began a period of expansion and in this period human brains have expanded almost 4x in size compared to all predecessors over 60 million years. Anthropologists theorize that the use of tools, complex social relationships, and expansion to climates not particularly well suited for human habitation forced the human brain to evolve in overdrive.

Now let’s come back to the topic at hand and the collection of technologies that are responsible for driving more innovation today than at any point in modern history. Machine learning.

We are witness to sophisticated machine learning frameworks that are dramatically improving the ability of machines to have human intelligence. Natural language interfaces are remarkably adept at understanding not just what we are saying but what we mean. Image technology is capable of replicating aspects of human vision with remarkable accuracy, along with speed and accuracy beyond human vision.

We see all of this come together with things like autonomous vehicles that develop behaviors that achieve not what is optimal performance but human performance. For example, the Google cars now cut a turn in a manner that replicates what human drivers naturally do. Human intelligence is being created in machine frameworks.

The next natural question to ask is whether or not machine learning can be used to accelerate machine learning. Much in the way that humans are interacting with other humans forced an expansion of our mental capabilities, leading to a machine-brain boom that follows the human evolution.

However, a machine with human intelligence and capacity still lacks something incredibly important about man-machine interfaces. Emotional capability.

I don’t believe we want machines to have emotions, for example, a robot that is sad or angry does not exactly get me excited. I believe that what we want machines to have is not an emotion at all, but rather the ability to interpret and adjust according to our emotions. Empathy, which is one of the most interesting of human capabilities and certainly not limited to humans as other animal species have demonstrated empathy in their social interactions.

In reading up about innovations in the automotive world, I learned of an interesting concept developed by major auto components supplier Faurecia. While not exactly what I envisioned, it is nonetheless an example of machines developing emotional intelligence that becomes a focus on the human interaction.

The result is the Active Wellness seating system, developed to improve the comfort and well-being of passengers. It’s the first system of its kind, detecting heart and breathing rates to determine an occupant’s stress level. Working in conjunction with a tablet inside the car, the seat can alert riders when it believes stress levels are too high, offering remedies such as therapeutic massage or increased seating ventilation.

Think of Siri having the capacity to adjust responses according to my mental state, sensing if I am sarcastic, happy, humorous, sad, or stressed. Go beyond the interaction style and consider the ability to alter responses based on emotional state. In this scenario, along with humanoid machine forms, I believe we can interact with machines on intellect and emotion.

To answer Lisa Padilla’s question: yes, I believe we will form emotional bonds with robots.

 

The Fascinating Implications for Autonomous Vehicles

I absolutely love intersecting public policy with technology trends, and if one thing is clear in my years of covering this it is that public policy reacts rather than anticipates technology advances. Politicians and bureaucrats alike, no matter how well-meaning, rarely take into account the 2nd order consequences of technology innovation, which I define as those consequences that are not obvious but often have a bigger impact than the immediate ones.

AVAutonomous, also known as driverless, cars present just such a scenario to consider. It’s easy to reflect on the safety implications and how we regulate the human intervention aspects, but let’s go a level deeper and consider what could happen with the single biggest impact on transportation since the invention of the automobile itself.

  • Traffic laws: Today’s laws are designed to accommodate distinctly human behavior, such as the fact that roadway speeds are discounted to take into account the fact that people will drive faster than the posted limit. In fact, the legal speed for any public road is the speed which is safe for the conditions. Machines are not bound by the limits of humans and it is reasonable to suggest that posted speed limits should be eliminated in favor of a network system of dynamic speeds set by the vehicles themselves based on accepted safety standards.
  • Driver licenses: If the car is doing the driving and I’m merely a passenger, do I need to have a license? Okay, human intervention will be mandated in this early period, so maybe we should have a different class of license that does not grant the right to drive, but rather to intervene. Removing age and license restrictions introduces a different set of consequences related to ownership and registration/licensing. More on that later.
  • DUI: Do we need DUI laws for autonomous vehicles if the software rather than the person is doing all the work?
  • Parking: An autonomous vehicle could drop me off at work and return home rather than having me pay for parking. Combine that with fleet-based ride-hailing for short hops and parking lots along with street parking get radically reduced.
  • Tolls: If an autonomous vehicle doesn’t require a person, how can the bridge and road tolls be collected on traditional toll booth manned infrastructure?
  • Liability: Who assumes it? We’re handing over operational responsibility to the vehicle itself, therefore it would be natural to assume that liability also transfers to the manufacturer. No? Sure, they could attempt to transfer it via EULA but at the end of the day a bug is going to have serious consequences that go beyond just being inconvenient.
  • Traffic infractions: Similar to liability, if the vehicle does create an infraction, who gets the ticket?
  • License plates: Need them? The autonomous vehicle is inherently connected, why does it need a license plate if, presumably, it won’t get pulled over and the registration information could be transmitted to an authenticated requester electronically?
  • Configurable roadways: Networked vehicles present an amazing opportunity to convert a fixed and inflexible infrastructure to a dynamic and reconfigurable one. Heavy traffic going northbound on 280, take one of the southbound lanes and use it for northbound traffic for 30 minutes.
  • First responders: Autonomous vehicles will reduce accidents, which suggests we will need fewer first responders. A self-managed fleet of vehicles will not require traffic enforcement.
  • Shifting revenue: Take away towing fees because cars are not parking, traffic fines, DUI fines, etc., and a large chunk of revenue going to the government goes away. Of course, the necessary manpower used to enforce regulation also goes away so maybe it balances out.

As you can see, fascinating to consider the implications and yet public officials rarely even go there…

Operation Email Purge

I’ve been on a crusade against my gmail inbox and am happy to report deleting over 100k unread emails in just a month. 100k unread emails… something is wrong with the state of email marketing when that happens.

Gmail provides a bunch of really good tools for managing email:

  • Unsubscribe link: If the sender has an unsub link in the footer you will likely see a handy link that Gmail inserts in the header. Use it.
  • Filters: Wow, where to start? Is:unread is a favorite.
  • Search Terms: Searching on terms like “casino” and “viagra” yields a massive number of emails I never wanted. Think about terms you never use in conversation and you will be amazed at how prevalent they are in email. Search on names other than your own, in my case Judy, James, John, Jennifer Nolan all resulted in hundreds of unwanted emails. Search on punctuated forms of your email that you don’t use, e.g. j.nolan@gmail.com. Search on phrases like “media briefing” and “embargo” will result in thousands if you are on PR lists. “Webinar” is another solid term to search on.
  • Search on sender. By far the biggest offender in email is notification email. Fortunately, they are easy to identifier based on subject verbage, but also sender. For example, “from:noreply@youtube.com” will yield all notif emails from YouTube.
  • Block. If a sender isn’t behaving well on email, e.g. bestbuy.ca, then block them. Now, nothing gets through.

Now here’s a couple of annoying email behaviors that marketers should be aware of:

  • Unsubscribe takes 10 business days… it takes a company just seconds to subscribe me, it should not take them longer than seconds to unsubscribe me. This is annoying because it reminds me how antiquated your infrastructure is.
  • Long messages. If your email exceeds the display length for Gmail, don’t send it.message too long
  • No unsubscribe. For a reason I cannot explain, PR agencies in Europe and Australia don’t use marketing automation to send emails, while U.S. based agencies do. The result is that U.S. agencies have unsub while the others get blocked.
  • Multi-click unsubscribe. Systems like Mailchimp and Constant Contact have really well constructed unsubscribe features. One click, maybe two and I’m out. If you unsubscribe requires more than 2 clicks, entering your email, or confirming and then sending another message to me after I unsubscribed telling me that I unsubscribed… you blew it.

I used all of the above to pare down my inbox. It didn’t happen quickly but now I’m in a rhythm and just 15 minutes every morning cleans things up. I target deleting a couple hundred email each day, which is more than I get each day so my inbox continues to shrink.

Artist Series – Naomi Edelberg Janches

It’s been quite some time since I last wrote about an artist I like. Today I’d like to get back in the habit of writing about something other than tech, and the subject of my attention is Naomi Edelberg Janches.

I can’t say that I have ever had a particular focus on stained glass but a couple of years ago we were replacing the doors on our garage with traditional carriage house doors and needed something special for the mahogany swing out doors. I went to Custommade.com and posted a general description for what I wanted.

P1000538Naomi was the first artist to respond out of over 40 in total and was very agreeable to the style I was pursuing. However, when I visited her website I was instantly captivated by her unique style and after some consideration I wrote Naomi that I wanted to trust to her the project and whatever she shipped to me would be what got installed. I gave her dimensions and a rough cut on the color palette we were designing around. Over the months that she was working on this glass, she honored my wish to not send any progress pictures, and the first time I viewed the design was when I unpacked them.

The crates arrived and much to my disappointment one of the panels had 2 cracks in it. Naomi was very accommodating to repairing the cracked glass. We also expanded the project to include 2 additional pieces of glass for the entry door and sidelight, for 6 panels in total. A picture does not do these stained glass pieces justice, in person they come alive and appear to dance with the light.

After 2 years of working together, it’s safe to say that not only did we connect but we also became good friends. She joined her husband on a business trip to the Bay Area and we hosted them at our house for dinner. It was a lovely evening.

It is safe to say that I am enthusiastic about her work not only out of the admiration for her creative ability but also because I am so fond of her personally. But it is the work that graces our house and in the glass you can see the  inspiration of the cosmos and nature with free-flowing lines of glass, impeccable placement of each pane, and celestial bodies represented in the glass orbs. We’re fortunate to have her work in our house, check out her website and facebook page for more info.

You Want Disruption? Can You Handle it?

johnny cashI’m on a panel tomorrow at the Mobile Future Forward event in Seattle. The panel discussion is on disruption, which we met on Friday to discuss over a quick conference call.

I actually don’t think a lot about disruption, which is not to say I don’t look for it. I just try to avoid being disrupted because you never realize the extent to which it’s happening until it’s already happened. The corollary to that is you really can’t plan on making it happen either.

This is why I don’t think about disruption. The great case studies for disruption in our business are all a result of executing really well against an idea that was well conceived and not impaired by compromise. Disruption happens when you don’t have anything to lose, and in every case disruption is the result of bringing together extraordinary people unified by a sense of purpose. Put another way, disruption is not a strategy, it is a consequence.

Encumbent players in any market NEVER disrupt the market they are in. They have too much to lose by leading the way to a new normal and insist on compromise as a core value. You have all seen this happen a million times, new products that threaten the status quo are nipped and tucked in ways that take everything potentially special away from them. There is a civility in these efforts that precludes them from breaking out… they may be successful but they won’t be disruptive. Transformation is not disruption.

You want disruption in your business? Hire the right fucking people! Search out combative, difficult, and argumentative people who care in a very personal way about the vision and purpose of the effort. Teams are messy because people themselves are messy, don’t risk a mediocre outcome by hiring people who excuse mediocrity and have toned down their passion in order to fit in. Hire people who won’t compromise.

Not compromising is not the same as not changing your answer. Wendy Lea once gave me an unintended compliment that I cherish to this day when she said “you are so damn stubborn… (and after awkward moment of silence) but I’ve seen you change your mind on big things.” Of course, if you get new information that supports a better direction than you originally pursued, change your mind. There are no prizes for who is the purest and most dogmatic, therefore changing your position in support of the common purpose is not only practical, it’s being smart.

Here’s what I’m looking for in the people I want to hire:

  1. They take things personally: Success and failure isn’t a clinical, sterile outcome. It involves emotions and sense of caring that goes beyond a job.
  2. Relentlessly curious: Want to infuse new thinking and approaches in any effort? Seek out knowledge in the unlikely of places, talk to people who have already done it, and don’t be afraid to share the random.
  3. Direct and confrontational: Don’t mix words, say what you think and don’t waste anyone’s time by waiting. If something is borked, call it out! If something is working, shout it out and do more of it.
  4. Laugh at themselves: Humor is a powerful antitode for tension, even more so when it is directed at yourself. This ain’t bean ball and even though it’s a tough, high stakes business that asks a lot of people, we can still have fun doing it.
  5. All the usual stuff: Smart, experienced, hungry… yeah, all that, but the inescapable fact is that there are a lot fo smart, experienced, and hungry people. These three attibutes by themselves are no indicator of future success.

 

Fingerprint Technology is the Next Privacy Catastrophe

OPMNew fallout today from the gift that keeps on giving, the Office of Personnel Management (OPM) hack. The news reports on this have focused on the standard PII elements along with the salacious possibilities associated with the disclosure information that is collected for security clearance applications.

 

FingerprintAn angle that has not been widely covered is the initial disclosure that 1.1 million fingerprints were also hacked. Today it is being reported that OPM has increased that number to 5.6 million fingerprints.

The nearly universal response to suggestions that people could be at risk is that the fingerprints are encrypted. Fair point, they are.

According to OPM, “federal experts believe that, as of now, the ability to misuse fingerprint data is limited.” The office acknowledged, however, that future technologies could take advantage of this information.

The government also said salt and fat were bad, and healthcare costs would go down.

Coincidentally, the NSA put out an advisory last month on Suite B elliptic curve cryptography that is widely used in the government, and is suitable for general national security use. Unlike Suite A, Suite B is widely used and available as a public standard.

According to the NSA, Suite B cryptography is not capable of withstanding advances in quantum computing.

Until this new suite (to replace Suite B) is developed and products are available implementing the quantum resistant suite, we will rely on current algorithms. For those partners and vendors that have not yet made the transition to Suite B elliptic curve algorithms, we recommend not making a significant expenditure to do so at this point but instead to prepare for the upcoming quantum resistant algorithm transition.

Well, this is reassuring… but let’s get back to the issue of fingerprint biometrics. The problem goes to the very nature of the biometric attribute itself, it is literally something about you and it is immutable. When someone hacks your fingerprints they have them forever. Forever.

I do have a horse in this race, having recently joined a speech biometrics company. Active speech verification has vulnerabilities, clearly, but one advantage over competing biometric technologies. In the event of a data breach that gives hackers the voice model data, an organization can simply force a re-enrollment for the participants and the integrity of the system is maintained. It’s the equivalent of forcing a password reset for your voice.

No system is without some vulnerability, but a system that does not provide for a reset capability is one that I have serious reservations about. With Apple TouchID and the upcoming Android M release with fingerprint support, fingerprint technology is mainstreaming. We are entering a period where fingerprint biometric data volume will explode and become an attractive target for hackers.

We’re building a speech verification and authentication service for developers who want to build speech biometrics into their apps using simple and reliable APIs. Sign up for news and launch updates, as well as early access, at knurld.io.

The Uncontrolled Implosion at VW

volkswagen_logo_bleeding_by_greenbob1986VW has been embroiled in a massive controversy. Here is the summary:

  1. VW has been marketing “clean diesel” technology on the basis of being environmentally friendly while also being fun to drive.
  2. Independent testing done by West Virginia University revealed that VW diesel cars were not in fact clean when compared to competitors, and exceeded EPA regulations governing emissions.
  3. VW asserted that the tests were inaccurate and offered to perform voluntary recalls to address specific issues.
  4. The EPA threatened to withhold certification for 2016 models – meaning VW would not be able to sell them.
  5. The company then admitted that they had engineered a defeat device which detected when the vehicle was undergoing emissions testing. In normal driving, the required emissions equipment was turned off.

The EPA has threatened the company with $18b in fines, which won’t come to fruition. BP paid just a 1/3 of that for their massive environmental disaster in the Gulf. It is hard to see how VW would be subject to a fine that equates to $37,344 per vehicle affected.

Samsung

The damage to the VW brand is incalculable right now, but I predict it will be a death blow to them in the U.S. market. Their problem is twofold, the first is that their market share here has been stuck in perpetual single digits across categories. With the U.S. being the most competitive car market globally, every point of share comes at great cost.

The second problem is that they are now on record admitting to a conspiracy to deceive regulators and consumers alike. They have been marketing a clean diesel product that they knew was not, and worse, had engineered components in the vehicles themselves to perpetuate that deception.

This scandal is spreading, fast. The company has already admitted that they know 11 million vehicles globally have this defeat mechanism installed. Countries are opening up their own investigations daily, the damage to the brand is no longer contained the U.S.

For the foreseeable future every news story about VW will be about a scandal, every car review, even in the bought off automotive media, will be compelled to append articles about VW cars, and current customers will have to navigate recalls and sarcastic comments about their choice of vehicle.

What reason does anyone have to buy a VW in light of these revelations? While not alone in the annals of recent car company scandals, the VW one is the most brazen in concept. Toyota and GM have both suffered scandals of incompetence with air bags and ignition switches respectfully, and in all fairness a lot of people died as a result of those failures, which is not the case with VW. Toyota and GM also have market leader positions while VW is a third tier player in the U.S. market, which doesn’t provide the company with much inertia to ride this out.

I think they are done. Call the moving company and buy your tickets back to Wolfsburg. At least they still have Audi and, thus far, they have not been implicated in this scandal.

Apple Shutters Hopstop

Hopstop was a revelation to me, a legitimate ah-ha moment. The light bulb brightly shining in front of me in the form of a smartphone.

We take for granted today that Google Maps can tell us how to get anywhere, and for a generation not familiar with paper maps, that was a pretty big deal. It wasn’t always like this and through the development phase of this technology it remained car-centric.

HT_hopstop_nt_130719_16x9_992I was going to NYC quite a bit in 2005 and finally decided to ditch the street for the subway, but the NYC subway system can be intimidating to those not accustomed to the numbering, schedules, and transfers. Hopstop helped me make the subway a regular part of my NYC experience and in the process showed me how apps welded to smartphones with implicit geolocation capabilities could be life altering.

Apple acquired Hopstop in 2013 and announced today that the service will shutdown next month.

Apple and Google are pitched in a heated battle for your eyeballs on their maps. We’ll see more services built into their respective mapping platforms, but I’m mixed in my opinion of whether more is better when it comes to maps. The usability of maps is very often a function of the degree to which they are de-crapified. Google Maps has crammed more stuff into the basic mapping functionality and the app hasn’t become more usable as a result when it comes to the singular purpose of getting me from point A-to-B.

I appreciate how much more it does for me, and continually find new ways to use it, but Google Maps has a lot of crap that clutters up the UI. Apple Maps will no doubt suffer a similar fate as Apple races to catch up to Google in the everything-and-the-kitchen-sink phase that maps are currently in.

Hopstop will forever be held in high regard for me. It achieved a rate accomplishment of doing something far better than any competing approach while also changing my life in a good way that reshaped my expectation of all that followed.

#LifeOnMars

I was listening to a media overview of the Life On Mars Project. The short version is that 6 people, 3 men and 3 women, will spend 365 in a biosphere located on the slopes of the Mauna Loa volcano in Hawaii. They will be permitted to leave the dome but will be required to wear a space suit to simulate the conditions on Mars.

In an unfortunate naming twist, the acronym for the effort is HI-SEAS. I think they should have gone with a different acronym.

The project will study the effects of long-term isolation and co-habitation. The results of the 1-year study will be used in the planning of an actual manned mission to Mars. This is certainly interesting research and the kind of work that can’t be simulated given the complexity of the human condition.

I thought it was curious that the crew features 3 men and 3 women. Given the fact that this is a 1-year project in close quarter conditions, with no external human contact, it would appear that the organizers were anticipating the human needs beyond food, shelter, clothing.

This leads to a number of interesting questions about the consequences of limited human interactions and self-enforced rules of order. What if one crewmember commits a crime against another crewmember? In a 36-foot diameter dome do you quarantine that person? Each crewmember holds specific skills that are necessary for the ongoing support of the station, putting an individual in isolation would deprive the rest of the crew of essential support.

What if someone unexpectedly binges and eats all the chocolate? Seems like a small thing but human conflict is often the result of small things piling up over time.

What happens when someone dies? This being a research project, one can presume that there would be an intervention in the event of life-threatening illness, but in space that isn’t possible so what would happen? Would you shoot the body out of a port, say a few words and move on? How do you replace the essential skills that person held?

There are obvious questions about long-term co-habitation in a small space that this research will shed light on. Entertainment, fitness, conflict resolution, communication, and mental health are all obvious questions, but it would be fascinating to learn about all the not so obvious issues that the planners have on their list.

I hope NASA departs from the usual media strategy of sunshine and rainbows to describe their work. This is a rare opportunity to shed light on the complexity of the human condition in anticipation of actual long-term cohabitation in space. I am also left to wonder why the International Space Station is not being used for this research given that it is an actual installation in outer space designed to support long-term co-habitation.

More on this topic (What's this?)
Market Outlook
Inteliquent Lands a Big One
Read more on Hong KONG&CHINA Gas at Wikinvest

The Ashley Madison Data Dump

madisonThe hack itself is an old story; Ashley Madison was hacked and profile information stolen. What is new is that the group behind the hack, Impact Team, dumped all the data. It has now been made available by many legitimate folks who created searchable online services against the data.

As much as I could enjoy the schadenfreude in this story, I simply cannot. I’m compelled to point out that just because an email is in the database does not mean it is a legitimate account. Email is a notoriously weak verified identifier, and while many websites have a sign-up flow for email verification, many don’t do anything to purge unverified emails. My thesis is that even unverified emails harvested in the sign-up flow have value for marketing purposes and, therefore, remain in the company’s database.

My email, jnolan@gmail.com, has been used by far too many “jnolans” to count. Often signing up for mundane services like car buying sites, but also for things that would certainly make my wife ask questions, like BlackPeopleFinder.com and an unrelated service for making arrangements with dominatrixes for a variety of, well, services.

The amount of crap I get from websites I have never visited is simple extraordinary. My oldest son has a gmail address that is first name only and I purge over a thousand emails from his account each month, and he’s only used it a few times for sending schoolwork.

2013101108The Impact Team has shrewdly wrapped themselves in a veil of moral righteousness to conceal a criminal act. While they aren’t stoning adulterers in the town square – or beheading them in a stadium – they are stealing personal information and using that in a form of extortion.

I find the entire affair, no pun intended, reprehensible and while AshleyMadison is itself objectionable, they are also a victim (of stupidity first and foremost). Despite complete awareness of the risks to the company and their customers, they did not employ best practices to secure their data. In addition to that, they had a sign-up flow and password recovery process that made it exceptionally easy to determine whether or not an email was in their user database. The flawed password recovery feature allowed for an entirely different line of attack employing social engineering to hijack individual accounts.

I won’t shed any tears if they shut down, which they likely will because recovering now is all but impossible, but I won’t celebrate the fact that a group of hackers brought their demise. To do so would welcome a global online sharia law where only those services that pass a moral test can exist.

PS- yeah, I searched for my email in the database! Who wouldn’t?