Duplex shows Google failing at ethical and creative AI design

Google CEO Sundar Pichai milked the woo from a clappy, home-turf developer mob at its I/ O conference in Mountain View this week with a demo of an in-the-works voice deputy feature that is conducive to the AI to make phone calls on behalf of its human owner.

The so-called’ Duplex’ feature of the Google Assistant was shown calling a hair salon to volume a woman’s hair cut, and ringing a eatery to try to book a table — merely to be told it did not accept bookings for less than five people.

At which point the AI changed tack and asked about wait times, earning its owner and controller, Google, the reassuring intel that there wouldn’t be a long wait at the elected time. Job done.

The voice system deployed human-sounding vocal cues, such as’ ums’ and’ ahs’ — to make the” conversational experience more comfy“, as Google couches it in a blog about its aims for the tech.

The voices Google used for the AI in the demos were not synthesized robotic tones but distinctly human-sounding, in both the female and male flavors it showcased.

Indeed, the AI pantomime was apparently realistic enough to convince some of the genuine humans on the other objective of the line that they were speaking to people.

At one point the bot’s’ mm-hmm’ answer even described appreciative giggles from a techie audience that clearly felt in on the’ joke’.

But while the home mob cheered enthusiastically at how capable Google had apparently constructed its prototype robot caller — with Pichai going on to sketch a grand vision of the AI saving people and industries day — the episode is worryingly suggestive of a company that views ethics as an after-the-fact consideration.

One it does not allow to trouble the trajectory of its engineering ingenuity.

A consideration which only seems to get a look in years into the AI dev process, at the cusp of a real-world rollout — which Pichai said would be coming shortly.

Deception by design

” Google’s experimentations do appear to have been designed to deceive ,” concurred Dr Thomas King, a researcher at the Oxford Internet Institute’s Digital Ethic Lab, discussing the Duplex demo.” Because their main hypothesis was’ can you distinguish this from a real person ?’. In this case it’s unclear why their hypothesis was about misrepresentation and not the user experience … You don’t inevitably need to deceive someone to give them a better user experience by voicing naturally. And if they had instead tested the hypothesis’ is this technology better than preceding versions or just as good as a human caller’ they would not have had to deceive people in the experiment.

” As for whether the technology itself is deceptive, I can’t really say what their intent is — but … even if they don’t intend it to delude you can say they’ve been negligent in not inducing sure it doesn’t deceive … So I can’t say it’s definitely deceptive, but there should be some kind of mechanism there to let people know what it is they are speaking to .”

” I’m at colleges and universities and if you’re going to do something which involves deception you have to really demonstrate there’s a scientific value in doing this ,” he added, agreeing that, as a general principle, humen should always be able to know that an AI they’re interacting with is not a person.

Because who — or what — you’re interacting with” shapes how we interact”, as he set it.” And if you start blurring the lines … then this can sew mistrust into all kinds of interactions — where we would become more suspicious as well as needlessly replacing people with meaningless agents .”

No such ethical dialogues troubled the I/ O stage, however.

Yet Pichai told Google had been working on the Duplex technology for “many years”, and went in so far as to claim the AI can ” understand the nuances of conversation” — albeit still plainly in very narrow scenarios, such as booking an appointment or reserving a table or asking a business for its opening hours on a specific date.

” It brings together all our investments over the years in natural language appreciation, deep learning, text to speech ,” he said.

What was yawningly absent from that list, and seemingly also lacking from the design of the tricksy Duplex experiment, was any sense that Google has a deep and nuanced appreciation of the ethical concerns at play around AI technologies that are powerful and capable enough of passing off as human — thereby playing lots of real people in the process.

The Duplex demos were pre-recorded, rather than live telephone call, but Pichai described the calls as “real” — indicating Google representatives had not in fact called the businesses ahead of time to warn them its robots might be calling in.

” We have many of these instances where the calls quite don’t go as expected but our assistant understands the context, the subtlety … and managed the interaction gracefully ,” he added after airing the restaurant unable-to-book example.

So Google appears to have trained Duplex to be robustly deceptive — i.e. to be able to reroute around derailed conversational expectations and still pass itself off as human — specific features Pichai lauded as’ graceful’.

And even if the AI’s performance was more patchy in the wild than Google’s demo suggested it’s clearly the CEO’s goal for the tech.

While trickster AIs might bring to mind the iconic Turing Test — where chatbot developers compete to develop conversational software capable of convincing human judges it’s not artificial — it should not.

Because the application of the Duplex technology does not sit within the context of a high profile and so clear competition. Nor was there a set of rules that everyone was shown and agreed to beforehand( at least so far as we are aware — if there were any rules Google wasn’t publicizing them ). Rather it seems to have unleashed the AI onto unsuspecting business staff who were just going about their day chores. Can you insure the ethical disconnect?

” The Turing Test has come to be a bellwether of testing whether your AI software is good or not, based on whether you can tell it apart from a human being ,” is King’s suggestion on why Google might have chosen a similar trick as an experimental showcase for Duplex.

” It’s very easy to tell seem how great our software is, people cannot tell it apart from a real human being — and perhaps that’s a much stronger selling point than if you say 90% of users favor this software to the previous software ,” he posits.” Facebook does A/ B testing but that’s probably less exciting — it’s not going to wow anyone to tell well consumers favor this slightly deeper shade of blue to a lighter shade of blue .”

Had Duplex been deployed within Turing Test conditions, King also induces the point that it’s rather less likely it would have taken in so many people — because, well, those slightly jarringly day ums and ahs would soon have been spotted, uncanny valley style.

Ergo, Google’s PR flavored’ AI test’ for Duplex is also rigged in its favor — to further supercharge a one-way promotional marketing message around artificial intelligence. So, in other words, say hello to yet another layer of fakery.

How could Google introduce Duplex in a way that would be ethical? King reckons it would need to state up front that it’s a robot and/ or use an appropriately synthetic voice so it’s instantly clear to anyone picking up the phone the caller is not human.

” If you were to use a robotic voice there would also be less of a risk that all of your voices that you’re synthesizing merely represent a small minority of the population spoke of’ BBC English’ and so, perhaps in a sense, use a robotic voice would even be less biased as well ,” he adds.

And of course , not being up front that Duplex is artificial embeds all sorts of other knock-on hazards, as King explained.

” If it’s not obvious that it’s a robot voice there’s a risk that people come to expect that most of these phone calls are not genuine. Now experiments have shown that many people do interact with AI software that is conversational just as they would another person but at the same period there is also evidence showing that some people do the exact opposite — and they become a lot ruder. Sometimes even abusive towards conversational software. So if you’re constantly interacting with these bots you’re not going to be as polite, perhaps, as you usually would, and who are likely have effects for when you get a genuine caller that you do not know is real or not. Or even if you know they’re real perhaps the route you interact with people has changed a bit .”

Safe to say, as autonomous systems get more powerful and capable of performing tasks that we would normally expect a human to be doing, the ethical considerations around those systems scale as exponentially large as the health risks applications. We’re really just getting started.

But if the world’s biggest and most powerful AI developers believe it’s totally fine to set ethics on the backburner then risks are going to spiraling up and out and things could go very badly indeed.

We’ve seen, for example, how microtargeted advertising platforms have been hijacked at scale by would-be election fiddlers. But the overarching hazard where AI and automation technologies are concerned is that humans become second class citizens vs the tools that are being claimed to be here to help us.

Pichai said the first — and still, as he set it, experimental — use of Duplex will be to supplement Google’s search services by filling in information about industries’ opening periods during periods when hours might inconveniently vary, such as public holidays.

Though for a company on a general mission to’ organize the world’s information and make it universally accessible and useful’ what’s to stop Google from — down the line — deploying vast phalanx of phone bots to ring and ask humen( and their associated businesses and organizations) for all sorts of expertise which the company can then liberally extract and inject into its multitude of connected services — monetizing the freebie human-augmented intel via our extra-engaged attention and the ads it serves alongside?

During the course of writing this article we reached out to Google’s press line several times to ask to discuss the ethics of Duplex with a relevant company representative. But ironically — or perhaps fittingly enough — our hand-typed emails received only automated responses.

Pichai did emphasize that the technology is still in development, and told Google wants to” work hard to get this right, get the user experience and the high expectations right for both businesses and users “.

But that’s still ethics as a tacked on afterthought — not where it should be: Locked in place as the keystone of AI system design.

And this at a time when platform-fueled AI problems, such as algorithmically fenced fake news, have snowballed into huge and ugly global scandals with very far reaching societal implications indeed — be it election interference or ethnic violence.

You really have to wonder what it would take to shake the’ first transgres it, later fix it’ethos of some of the tech industry’s major players…

Ethical guidance relating to what Google is doing here with the Duplex AI is actually pretty clear if you bother to read it — to the phase where even legislators are agreed on foundational basics, such as that AI needs to operate on “principles of intelligibility and fairness”, to borrow phrasing from just one of several political reports that have been published on the topic in recent years.

In short, misrepresentation is not cool. Not in humans. And absolutely not in the AIs that are supposed to be helping us.

Transparency as AI standard

The IEEE technical professional association put under a first draft of a framework to guide ethically designed AI systems at the back end of 2016 — which included guiding principles such as the need to ensure AI respects human rights, operates transparently and that automated decisions are accountable.

In the same year the UK’s BSI criteria body developed a specific criterion — BS 8611 Ethics design and application robots — which explicitly names identity deception( intentional or unintentional) as a societal risk, and warns that this approach will eventually erode trust in the technology.

” Avoid misrepresentation due to the behaviour and/ or appearance of the robot and is transparency of robotic nature ,” the BSI’s criterion advises.

It also warns against anthropomorphization due to the associated danger of misinterpretation — so Duplex’s ums and ahs don’t only suck because they’re fake but because they are misinforming and so deceptive, and also therefore carry the knock-on danger of undermining people’s trust in your service but also more widely still, in other people generally.

” Avoid unnecessary anthropomorphization ,” is the standard’s general guidance, with the further steer that the technique be reserved” merely for well-defined, limited and socially-accepted intents “.( Tricking workers into remotely conversing with robots probably wasn’t what they were thinking of .)

The standard also urges” clarification of intent to simulate human or not, or intended or expected behaviour “. So, yet again, don’t try and pass your bot off as human; you will be required to make it really clear it’s a robot.

For Duplex the transparency that Pichai said Google now intends to think about, at this late stage in the AI development process, would have been trivially easy to attains: It could just have programmed the assistant to tell up front:’ Hi, I’m a robot calling on behalf of Google — are you happy to talk to me ?’

Instead, Google chose to prioritize a demo’ wow’ factor — of proving Duplex pulling the woolen over busy and trusting humans’ eyes — and by doing so proved itself tonedeaf on the topic of ethical AI design.

Not a good look for Google. Nor indeed a good outlook for the rest of us who are subject to the algorithmic whims of tech giants as they flick the control switches on their society-sized platforms.

” As the development of AI systems grows and more research is carried out, it is important that ethical hazards associated with their utilize are highlighted and considered as part of the design ,” Dan Palmer, head of manufacturing at BSI, told us.” BS 8611 was developed … alongside scientists, academics, ethicists, philosophers and users. It explains that any autonomous system or robot should be accountable, truthful and unprejudiced.

” The standard creates a number of potential ethical hazards that are relevant to the Google Duplex; one of these is the risk of AI machines becoming sexist or racist due to a biased data feed. This surfaced prominently when Twitter users influenced Microsoft’s AI chatbot, Tay, to spew out offensive messages.

” Another contentious topic is whether forming an emotional bond with a robot is desirable, especially if the voice assistant interacts with the elderly or children. Other guidelines on new hazards that should be considered include: robot misrepresentation, robot addiction and the potential for a learning system to surpass its remit.

” Ultimately, it must always be transparent who is responsible for the behavior of any voice assistant or robot, even if it behaves autonomously .”

Yet despite all the thoughtful ethical guidance and research that’s already been produced, and is out there for the reading, here we are again being proven the same tired tech industry playbook praising engineering abilities in a shiny bubble, stripped of human context and societal consideration, and hung in front of an uncritical audience to see how loud they’ll cheer.

Leaving important matter — over the ethics of Google’s AI experimentations and also, more broadly, over the mainstream vision of AI assistance it’s so keenly trying to sell us — to hang and hang.

Questions like how much genuine utility there might be for the kinds of AI applications it’s telling us we’ll all want to use, even as it prepares to push these apps on us, because it can — as a consequence of its great platform power and reach.

A core’ uncanny valley-ish’ contradiction may explain Google’s choice of deception for its Duplex demo: Humen don’t inevitably like speaking to machines. Indeed, oftentimes they prefer to speak to other humans. It’s just more meaningful to have your existence registered by a fellow pulse-carrier. So if an AI exposes itself to be a robot the human who picked up the phone are most likely only put it straight back down again.

” Going back to the subterfuge, it’s fine if it’s replacing meaningless interactions but not if it’s intending to replace meaningful interactions ,” King told us.” So if it’s clear that it’s synthetic and you can’t inevitably use it in a context where people actually want a human to do that job. I think that’s the right approach to take.

” It matters not just that your hairdresser appears to be listening to you but that they are actually listening to you and that they are mirroring some of your emotions. And to replace that kind of work with something synthetic — I don’t think it stimulates much sense.

” But at the same time if you expose it’s synthetic it’s not likely to replace that kind of work .”

So actually Google’s Duplex sleight of hand may be trying to conceal the fact AIs won’t be able to replace as many human tasks as technologists like to think they will. Not unless lots of currently meaningful interactions are rendered meaningless. Which would be a massive human expense that societies would have to — at very least — debate long and hard.

Trying to avoid such a debate from taking place by pretending there’s nothing ethical to see here is, hopefully , not Google’s designed intention.

King also stimulates the point that the Duplex system is( at least for now) computationally costly.” Which means that Google cannot and should not only release this as software that anyone can run on their home computers.

” Which means they can also control how it is used, and in what context — and they can also is ensured will only be used with certain safeguards built in. So I believe the experiments are maybe not the best of signs but the real exam is likely to be how they release it — and will they construct the safeguards that people demand into the software ,” he adds.

As well as a lack of visible precautions in the Duplex demo, there’s also — I would argue — a curious absence of imagination on display.

Had Google been bold enough to disclose its robot interlocutor it might have guessed more about how it could have designed that experience to be both clearly not human but also fun or even funny. Suppose of how much life can be injected into animated cartoon characters, for example, which are very clearly not human yet are enormously popular because people find them entertaining and feel they come alive in their own way.

It actually attains you wonder whether, at some foundational level, Google absence trust in both what AI technology can do and “in ones own” creative abilities to breath new life into these emergent synthetic experiences.

Make sure to visit: CapGeneration.com


Josh.ai raises $11 million for a premium home automation system with a smarter AI

One of the promises of voice-based calculating is the ability to attain home automation simpler something that major tech companies, including Amazon, Apple and Google, are now tackling with their own voice assistants and smart speakers. But their solutions are still somewhat clunky, both in terms of the software interface for configuring your smart home and the voice commands you use to take actions. Thats where the startup Josh.aicomes in.

The company has now created $11 million to design a better voice-controlled system for smart homes, and will afterward this year release its own hardware dedicated to this purpose.

Headquartered in Denver with offices in L.A ., Josh.ai is the product of serial entrepreneursAlex Capecelatro, CEO, and Tim Gill, CTO. The two previously worked together on a social recommendations app Yeti, which had begun its life as At The Pool, andwas sold back in 2015. Gill, who had previously founded and sold Quark( Quark XPress ), had joined Yeti as a technical advisor, and wrote a number of the algorithm being implemented in the app.

Following the sale of Yeti, the two teamed up again to work on a project in the smart home space something they were both interested in for personal reasons.

Gill, for example, had spent years developing his own home automation system his version of Mark Zuckerbergs Jarvis to run inside the large residential property he was building in Denver.

He was well underway in building the house and understanding what the competition looked likewhat the product offerings looked like, explainsCapecelatro. And he was pretty dissatisfied with what was out there.

Meanwhile, Capecelatro was also constructing a home for himself in L.A ., and running into the same problems.

I was just amazed that all of the big automation systems Crestron, Control4, and Savant they cost hundreds of thousands of dollars, and the[ user interface] looks like its from the 90 s, he says. It was weird that for a ton of money in my home where you want to have a delightful experience, the best offerings on the table merely werent that good.

The founders insured a need in the market for something that sits above mass market solutions, like Apples Home app, or Alexas smart home control, which focus more on tying together after-market devices, like security cameras, smart doorbells, or smart illuminates like Philips Hue.

They founded the startup Josh.ai in March 2015, and shipped the first product the following year.

The solution, as it exists today, includes a kit with a Mac mini and iPad, and software that runs the home. After plugging in the Mac, Josh.ai auto-discovers devices on the network. It can identify those from over 50 manufacturers. For example, it can control lighting and shades like those from Lutron, music systems like Sonos, dozens of brands of security cameras, Nest thermostats, Samsung smart TVs, and even more niche products like Global Cachs box for controlling IR devices( such as your not-so-smart TVs ).

The automatic speech recognition( AKA speech-to-text) portion of Josh.ais system is handled in the cloud, while Mac mini manages the natural language processing to know what your commands mean.

What builds Josh.ai unique “wasnt just” its software interface, but how users interact with the system. You speak to the voice deputy Josh to tell the home what to do.( You can also change its name if thats an issue, or even pick from a variety of male and female voices and accents .)

Josh, or the wake word youve chosen, precedes your command, which can be spoken use more natural language. The system is better than many when it comes to construing what you entail, by nature of its single-purpose focus on home automation.

For instance, you can tell Josh to turn it off , and it will know what it entails because it remembers what it had turned on before. Or you can say, its hot in here , and Josh will know how to adjust your thermostat.

It can also deep-link to streaming video content, so you can ask to watch Planet Earth, and Josh will turn on the TV, switch to the right input, launch Netflix, then start playing the show.

Josh.ai supports scenes, as well, allowing you to configure a number of devices to work together like suns, shades, music, fans, thermostats, and other switches. That route, you can say things like turn everything off , and Josh knows to shut down all the connected devices in the home.

Where the system gets really smart is in its ability to handle complex, compound commands entailing controlling multiple devices in one sentence.

You can say to Josh, play Simon and Garfunkel and turn on the suns , for example. Or, play Explosions in the sky in the kitchen, and play Simon and Garfunkel in the living room . Other systems could get tripped up by the and and the in the in the artists names, but Josh.ai understands when those terms are a transgres between two commands, and when theyre part of something else.

The current system which was largely designed for high-end homes is sold by professional integrators at around $10,000 and up, depending on the components involved. To date, the team has sold more than 50 and fewer than 100 installations.

Josh.ai can work over your Echo or Google Home, if you prefer, and includes interfaces for iOS, Android and the web. But the company is now preparing to launch its own, farfield mic answer in a new hardware device thats built specifically for use in the home.

While the new hardware will perform some basic virtual assistant kind chores telling you the climate, perhaps( the company isnt corroborating specific features at this time) the main focus will be on home automation.

Above: a tease of the new device

The hardware wont be a cylindrical shape like Echo or Google Home, but will be designed with an aesthetic appeal in mind.

It also wont be super cheap.

It will still be a premium product, but it will be a lot less than where the present product is. And the idea is this will enable our mass market rollout in probably a year to eighteen months , notesCapecelatro, speaking of his plan to keep bringing Josh.ais technology to ever larger audiences.

Josh.ai, a team of 15 soon to be 25, lately closed on$ 8 million in new funding, largely from the founders personal networks. The investors names arent being disclosed because theyre not institutional firms. To date, Josh.ai has raised $11 million, but has not yet added anyone to its board.

Make sure to visit: CapGeneration.com