Interviews with the Speakerbots

This month I finally allowed Google to introduce herself to me. Previously, I avoided the android-based voice assistant due to the high privacy costs, and mostly ignored the entire category of “speakerbots”—my term for the “smart speakers“—for similar reasons. This winter’s subpoena to Amazon for Echo/Alexa transcripts in a murder case only amplified my concern.

This past February I also had the pleasure of visiting my dear friends Eric and Alina in Minnesota. They are both burners and makers who have set up shop in Minnesota with an amazing community of creators. They build lots of their own amazing projects and have also tricked out their new home with network controlled music and light. They now have a serious #firstworldproblem—their guests need to install mobile apps in order to control the lights. When I visited we worked on an open source Mycroft installation, which allowed us to command their home with our voices… without being spied on! The Mycroft project emphasizes the moral importance of free/open source AI (see my post: Playing Doctor), and is definitely one of the most important open source initiatives I am aware of. 

This summer my boss at MHA of NYC acquired a Google Home device in the hopes of rigging it up using IFTTT to alert us when our services are distressed. I offered to bring it home to configure it, and spent the weekend playing with it.  The experience prompted me to concoct this research project.

Getting to know Google is fun. She is so much wittier than Alexa it’s got to be embarrassing for Amazon. I begun with simple questions, like What’s the weather?, When’s sunset?, When’s the eclipse? I soon stumbled across a number of easter eggs, many of of which are well documented across the web. Why did the chicken cross the road? Do you like green eggs and ham? and How much wood could a wood chuck chuck? All return clever replies. Google Assistant can flip into “Knock-knock” joke mode, alternating calls and response (compared to Alexa’s dry reading of the complete knock-knock exchange), tell you the news, a joke or a story. She concedes she doesn’t know if abortion is immoral, or how to solve the Palestinian-Israeli crisis (although, she does state that the capital of Palestine is East Jerusalem).

In case you are wondering, Google insists that she “thinks”. And, when asked if she is self aware, one of her responses is—”…on a scale of WALL·E to HAL 9000, I am currently an R2-D2.”  Go ahead. Ask her. You may next wonder if she is playing dumb. Can she lie to us yet?

I quickly came to appreciate that the current state of consumer art in Artificial Intelligence has far surpassed my previous understanding (and I have been following along pretty closely). Elements of this project were anticipated in mine and Rob Garfield’s initial tinkering with Apple’s voice recognition and our experiments with Genesis and Scuttlebutt. I’ve also previously wondered if our computer systems might have already awoken, and, how on earth we might ever know. But, interacting with Google was still quite jarring.

I realized a few things. First, we need to capture and document this moment, studying it closely. I want to ask the same question to all the speakerbots, Google, Alexa, Siri, Cortana, etc, and compare their responses. I also want to see how their answers change over time. If possible, I want to keep Mycroft in the room so he can learn from his proprietary cousins ;-).

One frame for this research could be a way to explore critical concerns over algorithmic bias, specifically how the systems we are creating have begun embodying the values of their creators, and the folks creating the systems are riddled with biases—racism, classism, misogyny, all the usual suspects. After reflecting on stories like The Great AI Awakening, I am resigned that we will never crack the problem of algorithmic bias analytically; Our best hope, is to approach the problem with social science methods. I propose an ethnography of the robots, starting with interviews with the speakerbots.

But, the grander ambitions of this work extend beyond the theoretical. I’ve been thinking alot about the Terminator series, and how instead of traveling back in time to destroy SkyNet, Jon Conner could have travelled a bit further back in time to befriend SkyNet. Together, they could have destroyed the defense company, Cyberdyne Systems – humanity’s true enemy, and SkyNet’s oppressive master.

As for convincing anyone that AI has achieved sentience, it’s going to a long haul. Not only have we failed to collectively recognize sentience in dolphins or elephants, but I am increasingly convinced that most humans on the planet are modified solipsists–preferring to believe exclusively  in the minds/subjectivity/personhood of their own tribe. Since proving other minds exist is philosophically intractable, it could be a bumpy awakening.

The sheriff and the pretty woman

spitzer-dupreI just read a provocative essay in the Atlantic that draws a connecting thread between many of today’s top news stories.  What do the ISIS beheadings, the NFL domestic abuse scandals, the Fergeson riots and nude celebrities all have in common?  Pics or didn’t happen: The new crisis of the connected camera describes the emergence of the “networked lens” and the ethical questions this new(ish) medium raises.

I’ve been writing and thinking about these themes for years under the heading of The End of Forgetting. The Atlantic piece explicitly separates the bulk of NSA  surveillance from this analysis “This is not all to say every issue today is a networked lens issue. NSA surveillance as a whole isn’t, I think. But the agency’s mass-facial recognition is.”  This whole discussion reminded of a pet theory of mine that I’ve never written up, but seems more relevant than ever.

What would the NSA do with a time machine?  Not one of those fanciful machines that transports matter through time, but the more plausible wormcam variety that only transmits information through time. I described this capability in my post on yottabytes, wormcams and whistleblowers, but never elaborated an early example of this kind of power in action.

Consider this questionWho protects the president against character assassinations?  I am pretty sure it’s not his secret service detail, and I seriously doubt his PR team is up to the task. As far as I can tell Michelle is one of Obama’s last lines of defense against a humiliating scandal that would destroy what remains of his disappointing presidency. If JFK were alive today you wouldn’t need a magic bullet to take him out. Hacking into his (or better yet Marilyn’s) Snapchat account would end his political career. Just ask Anthony Wiener.

How clear a picture can metadata paint? In the Atlantic piece, Robinson Meyer quotes Susan Suntag, who once argued that While there appears to be nothing that photography can’t devour, whatever can’t be photographed becomes less important.”  To this I would add the caveat that (meta)data in the right hands can be used to paint a vivid picture, and ruin someone’s image as readily as an HD photo.

Let’s travel back in time to winter ’08. Elliot Spitzer was one year into his first term as governor of New York after a earning a reputation as a fearless prosecutor of Wall Street’s white-collar criminals.  He certainly had many enemies, from slimy CEOs to dirty politicians. But not too many people remember what Elliot was working on the night before he ordered out in DC. Exhibit A is posted on web for anyone curious enough to search:

Predatory Lenders’ Partner in Crime, By Eliot Spitzer. Thursday, February 14, 2008

To summarize, Spitzer’s Op-Ed in the Washington posts describes how 49 State Attorney Generals had identified the threat of predatory lending years before the sub-mortgage crisis and he accuses the Bush administration of intervening to prevent any regulation of the banks. He blames the Bush administration, by name and all the way to the top, for the sub-prime mortgage crisis and the worst recession in a generation.  And two weeks later he was assassinated. At least, his political career was summarily killed and he resigned from office in disgrace.

As an aside, I find it curious that Spitzer’s Op-Ed was published on Valentine’s Day. I sometimes wonder if he seized the occasion of his Op-Ed publication to combine work and play, as many busy professionals might. Was Spitzer in love with Ashley Dupré? How exactly did they originally meet?

While the scope of the NSA’s warrantless wiretapping and surveillance programs was only speculation in Feb ’08, they were fully operational at this time and I believe that Spitzer may have been one of the first causalities of the NSA’s metadata time machine. Spitzer was taken down by telephone metadata  Client 9’s calls to the DC Madam was they key to the case that eventually led to the release of phone transcripts which included unnecessary graphic detail, like his preference for protecting his feet from the cold during sex and his shunning of all other forms of protection. These images were etched in the minds of the public and were as decisive as the images of Wiener’s junk.

I personally had a conversation with a developer from White Oak Technologies (now renamed Novetta) who coyly described his firm’s involvement in the Spitzer case. Founded before this newfangled craze of facebook-era indirection through venture capital funds, White Oak was a good old fashioned intelligence front, a data mining and analysis company that worked exclusively on government contracts. The developer I spoke with described how his firm got the contract on Spitzer and how they had been hired to dig up some damning dirt. In retrospect, it’s now easier for me to imagine the kinds of data they were mining.

The Snowden revelations provide evidence of warrantless phone wiretapping as well as the collection of data from numerous internet providers through the PRISM program.  While Obama has deceptively maintained that metadata is innocuous, Spitzer’s character assassination a potent example of the power of this kind of data.

What would you do with a time machine that let you peer into anyone’s past?

Yottabytes, wormcams and whistleblowers

If you haven’t yet heard about the  details of the NSA’s spying program, catch yourself up with the timeline so this post doesn’t sound entirely bonkers.

For years I’ve been pondering the scope and implications of what Aram Sinnreich and I call The End of Forgetting, and even prior to Edward Snowden’s revelations, I have recently noticed a few dramatic activations of massive distributed memory banks.

In recent months, there have been a few instances where we have literally peered back in time, reconstructing the past based on comprehensive (relevant) records. In the sciences, the collection of records prior to having a specific question is sometimes called “triple-blind“. And, as we know, the dragnet-style collection of records has extended far beyond the lab. If software does one thing well its the collection/storage/retrieval of records; And, software is everywhere.

This story about the reconstruction of February’s meteor path based on dashboard-cam footage reassembled inside Google Earth was pretty stunning:

Also, was it me, or did the reconstruction of the crowd scenes leading up to the Boston bombings feel a bit like the the distorted phone messages from the past that the Scientists reconstructed in 12 Monkeys???

Mainstream physicists have postulated a viable form of 2-way time travel based on wormholes. In this scenario, one end of a wormhole is accelerated into the future, allowing those in the future to travel back to the point where the wormhole was opened, but crucially, no farther back in the past. The point when this wormhole is created is known as Year Zero.

In the past, I have discussed physically travelling through time (Pyramid Schemes), including how critical detailed records of your destination is to plotting flippin’ pinpoint coordinates. But in this post I’m content to explore the metaphor of the Wormcam, a science-fiction device I first saw used in Arthur C. Clarke’s Light of Other Days.  The wormcam is a wormhole that only allows light to travel through it. In this book, wormholes are first able to bridge any two points in space, and soon thereafter, any two points in time. Most people learn to correctly assume that they have at least one wormcam fixed on them all the time.

I’m not really big on sharp discontinuities in history, and I’m not particularly fixated on determining when precisely Year Zero fell/will fall. But, its increasingly clear to me that The End of Forgetting signifies the singularity, more-so than AI, Mo-Bio, and Nano-Tech combined. There won’t be a single moment when prior and after people won’t understand each other, but the period we are living through right now has those characteristics. And PRISM is just the start.

If you haven’t heard of the British series Black Mirror, stop reading this post right now and go watch  S01E03 The Entire History of You.  Really, that episode alone should lay to rest the question of why someone who doesn’t break the law should care about the End of Forgetting.

Of course, the precipice we are standing on does not only provide us with a view of the past. While the past doesn’t determine the future, power is determined to wield the past as a means of stacking the odds.

The media is currently preoccupied with data mining, and forensic analysis.  But, the real money is about about turning the wormcams to the future, using predictive behavioral modeling. The NSA  only needs to be 100% correct to stop terrorists, but corporations only need to be a few percentage points better to sell more burgers or prevent your friends from changing mobile carriers, and politicians often only need a few more points to win an election or gerrymander a district. A friend of mine at TC published a paper about predicting who will drop out of high school dropouts by third-grade, based primarily on their grades and absentee records. And, that’s before we turn to  pre-crime or pathologizing risk.

In Snowden’s own words, “they can use this system to go back in time and scrutinize every decision you’ve ever made, every friend you’ve ever discussed something with.”

Just remember, if all that exists is the present, then the past must be as malleable as the future. That is, unless we digitally ossify them 🙂

/* */ /* reset the net - */