FCC slaps robocaller with record $120M fine, but it’s like ’emptying the ocean with a teaspoon’

Whoever thought we would leave telemarketing behind in this brave new smartphone world of ours lacked imagination. Robocalls are a menace growing in volume and even a massive $120 million fine leveled against a prominent source of them by the FCC likely won’t stem the flood.

The fine was announced today during the FCC’s monthly open meeting: a Mr Adrian Abramovich was responsible for nearly 100 million robocalls over a three-month period, and will almost certainly be bankrupted by this record forfteiture.

“Our decision sends a loud and clear message,” said FCC Chairman Ajit Pai in a statement. “This FCC is an active cop on the beat and will throw the book at anyone who violates our spoofing and robocall rules and harms consumers.”

That sounds impressive until you hear that these calls took place in 2016, and meanwhile there were 3.4 billion robocalls made last month alone. Commissioner Jessica Rosenworcel applauds the fine, but questions the practicality of pursuing damages when actions need to be taken to prevent the crimes in the first place.

“Let’s be honest,” she wrote in a statement, “Going after a single bad actor is emptying the ocean with a teaspoon.”

She points out that a set of rules designed to prevent robocalls was overturned a couple months ago, and that 20 petitions to the FCC under those rules for legal exemptions and such have yet to be addressed. And a technology designed to prevent robocalls altogether, recommended in a report more than a year ago and currently set to be implemented in Canada in 2019, has no such date here in the States.

As someone who gets these robocalls all the time, I fully support both this fine and the more serious measures Rosenworcel suggests. And the faster the better, I literally got one while writing this story.

The many twists and turns of hardware

Note: This is the final article in a three-part series on valuation thoughts for common sectors of venture-capital investment. The first article, which attempts to make sense of the SaaS revenue multiple, can be found here; the second, on public marketplaces can be found here.

Over the past year, the VC-backed hardware category got a big boost — Roku was the best-performing tech IPO of 2017 and Ring was acquired by Amazon for a price rumored to exceed $1 billion. In addition to selling into large, strategic markets, both companies have excellent business models. Ring sells a high-margin subscription across a high percentage of its customer base and Roku successfully monetizes its 19 million users through ads and licensing fees.

In the context of these splashy exits, it is interesting to consider the key factors that have made for valuable hardware companies against a backdrop of an investment sector that has often been maligned through the years, as I’m sure we’ve all heard the trope that “hardware is hard.” Despite this perception, hardware investment has grown much faster than the overall VC market since 2010, as shown below.

Source: TechCrunch

A large part of this investment growth has to do with the fact that we’ve seen larger exits in hardware over the past few years than ever before. Starting with Dropcam’s* $555 million acquisition in 2014, we’ve seen a number of impressive outcomes in the category, from large acquisitions like Oculus ($2 billion), Beats ($3 billion) and Nest ($3.2 billion) to IPOs like GoPro ($1.2 billion), Fitbit ($3 billion) and Roku* ($1.3 billion)**. Unfortunately for the sector, a few of these companies have underperformed since exit; notably, GoPro and Fitbit have both cratered in the public markets. 

As of April 3, 2018, both stocks traded at less than 1x trailing revenue, a far cry from the multiples of forward revenue given to other tech companies. Roku, on the other hand, continues to perform as a stock market darling, trading at approximately 6x trailing revenue and a market cap of $3.1 billion. What sets them so far apart?

The simple answer is their business model — Roku generates a significant amount of high gross margin platform revenue, while GoPro and Fitbit are reliant on continued hardware sales to drive future business, a revenue stream that has been stagnant to declining. However, Roku’s platform is one successful hardware business model; in this article I’ll explore four others — Attach, Replacement, Razor and Blades and Chunk.


“Attaching” a high gross margin annuity stream from a subscription to a hardware sale is a goal for many hardware startups. However, this is often easier said than done — as it’s critical to nail the alignment of the subscription service to the core value proposition of the hardware.

For example, Fitbit rolled out coaching, but people buy Fitbit to track activity and sleep — and this mismatch resulted in a low attach rate. On the other hand, Ring’s subscription allows users to view past doorbell activity, which aligns perfectly with customers looking to improve home security. Similarly, Dropcam sold a subscription for video storage, and at an approximate 40 percent attach rate created a strong economic model. Generally, we’ve found that the attach rate necessary to create a viable business should be at least in the 15-20 percent range.


Unlike the “Attach” business model that sells services directly related to improving the core functionality of the hardware device, “Platform” business models create ancillary revenue streams that materialize when users regularly engage with their hardware. I consider Roku or Apple to be in this category; by having us glued to our smartphones or TV screens, these companies earn the privilege of monetizing an app store or serving us targeted advertisements. Here, the revenue stream is not tied directly to the initial sale, and can conceivably scale well beyond the hardware margin that is generated.

In fact, AWS is one of the more successful recent examples of a hardware platform — by originally farming out the capacity from existing servers in use by the company, Amazon has generated an enormously profitable business, with more than $5 billion in quarterly revenue.


Despite the amazing economics of Apple’s App Store, as of the company’s latest quarterly earnings report, less than 10 percent of their nearly $80 billion in quarterly revenue came from the “Services” category, which includes their digital content and services such as the App Store.

What really drives value to Apple is the replacement rate of their core money-maker — the iPhone. With the average consumer upgrading their iPhone every two to three years, Apple creates a massive recurring revenue stream that continues to compound with growth in the install base. Contrast this with GoPro, where part of the reason for its poor market performance has been its inability to get customers to buy a new camera — once you have a camera that works “well enough” there is little incentive to come back for more.

Razor and Blades

The best example of this is Dollar Shave Club, which quite literally sold razors and blades on its way to a $1 billion acquisition by Unilever. This business model usually involves a low or zero gross margin sale on the initial “Razor” followed by a long-term recurring subscription of “Blades,” without which the original hardware product wouldn’t work. Recent venture examples include categories like 3D printers, but this model isn’t anything new — think of your coffee machine!


Is it still possible to build a large hardware business if you don’t have any of the recurring revenue models mentioned above? Yes — just try to make thousands of dollars in gross profit every time you sell something — like Tesla does. At 23 percent gross margin and an average selling price in the $100,000 range, you’d need more than a lifetime of iPhones to even approach one car’s worth of margin!

So, while I don’t think anyone would disagree that building a successful hardware business has quite literally many more moving parts than software, it’s interesting to consider the nuances of different hardware business models.

While it’s clear that in most cases, recurring revenue is king, it’s difficult to say that any of these models are intrinsically more superior, as large businesses have been built in each of the five categories covered above. However, if forced to choose, a “Platform” model seems to offer the most unbounded upside as it’s indicative of a higher engagement product and isn’t indexed to the original value of the product (some people certainly spend more on the App Store than on the iPhone purchase).

While it’s easy to take a narrow view of VC-hardware investing based on the outcome of a few splashy tech gadgets, broadening our aperture just a bit shows us that large hardware businesses have been built across a variety of industries and business models, and many more successes are yet to come.

*Indicates a Menlo Ventures investment

**Initial value at IPO

Google Pay’s app adds boarding passes, tickets, p2p payments and more

Google Pay got a big upgrade at Google I/O this week. At a breakout session, Google announced a series of changes to its payments platform, recently rebranded from Android Pay, including support for peer-to-peer payments in the main Google Pay app; online payments support in all browsers; the ability to see all payments in a single place, instead of just those in-store; and support for tickets and boarding passes in Google Pay’s APIs, among several other things.

Some of Google Pay’s expansions were previously announced, like its planned support for more browsers and devices, for example.

However, the company detailed a host of other features at I/O that are now rolling out across the Google Pay platform.

One notable addition is support for peer-to-peer payments which is being added to the Google Pay app in the U.S. and the U.K.

And that transaction history, along with users’ other payments, will all be consolidated into one place.

“In an upcoming update of the Google Pay app, we’re going to allow you to manage all the payment methods in your Google account – not just the payment methods that you used to pay in-store,” said Gerardo Capiel, Product Management lead at Google Pay, during the session at I/O. “And even better, we’re going to provide you with a holistic view of all your transactions – whether they be on Google apps and services, such as Play and YouTube, whether they be with third-party merchants, such as Walgreens and Uber, or whether they’re transactions you’ve made to friends and families via our peer-to-peer service,” he said.

The company also said it would allow users to send and request money, manage payment info linked to their Google accounts, and see their transaction history on the web with the Google Pay iOS app, too.

And because I/O is a developer conference, many of the new additions were in the form new and updated APIs.

For starters, Google launched a new API for incorporating Google Pay into other third-party apps.

“Via our APIs, we’re going to enable these ready-to-pay users [who already have payment information stored with Google Pay] to also checkout quickly and easily in your own apps and websites,” Capiel said.

The benefit to those developers who add Google Pay support is an increase in conversion rates and faster monetization, he noted.

Plus, Google added support for tickets and boarding passes to the Google Pay APIs, where they joined the existing support for offers and loyalty cards.

This allows companies such as Urban Airship or DotDashPay to help business clients distribute and update their passes and tickets to Google Pay users.

“It shows an even stronger commitment on Google Pay’s part to make the digital wallet a priority,” Sean Arietta, founder and CEO of DotDashPay, told TechCrunch, following the presentation. “It also reinforces their focus on partners like DotDashPay to help build connections between consumers and brands. The fact that they are specifically highlighting a complete experience that starts with payments and ends with an NFC tap-to-identify, is really powerful. It makes the Google Pay story now complete,” he added.

Urban Airship was also touting the changes earlier this week, via a press release.

“We help businesses reinvent the customer experience by delivering the right information at the right time on any digital channel, and mobile wallets fill an increasingly critical role in that vision,” Brett Caine, CEO and president of Urban Airship, said in a statement. “Google Pay’s new support for tickets and boarding passes means customers will always have up-to-date information when they need it most – on the go.”

Some of Google’s early access partners on ticketing include Singapore Airlines, Eventbrite, Southwest, and FortressGB, which handles major soccer league tickets in the U.K. and elsewhere.

In terms of transit-related announcements, Google added a few more partners who will soon adopt Google Pay integration, including Vancouver, Canada and the U.K. bus system, following recent launches in Las Vegas and Portland.

The company also offered an update on Google Pay’s traction, noting the Google Pay app just passed 100 million downloads in the Google Play store, where it’s available to users in 18 markets worldwide.

Soon, Google said it will launch many of the core features and the Google Pay app globally to billions of Google users worldwide.

AWS launches an undo feature for its Aurora database service

Aurora, AWS’s managed MySQL and PostgreSQL database service, is getting an undo feature. As the company announced today, the new Aurora Backtrack feature will allow developers to “turn back time.” For now, this only works for MySQL databases, though. Developers have to opt in to this feature and it only works for newly created database clusters or clusters that have been restored from backup.

The service does this by keeping a log of all transactions for a set amount of time (up to 72 hours). When things go bad after you dropped the wrong table in your production database, you simply pause your application and select the point in time that you want to go back to. Aurora will then pause the database, too, close all open connections and drop anything that hasn’t been committed yet, before rolling back to its state before the error occurred.

Being able to reverse transactions isn’t completely new, of course. Many a database system has implemented some version of this already, including MySQL, though they are often far more limited in scope compared to what AWS announced today.

As AWS Chief Evangelist Jeff Barr notes in today’s announcement, disaster recovery isn’t the only use case here. “I’m sure you can think of some creative and non-obvious use cases for this cool new feature,” he writes. “For example, you could use it to restore a test database after running a test that makes changes to the database. You can initiate the restoration from the API or the CLI, making it easy to integrate into your existing test framework.”

Aurora Backtrack is now available to all developers. It will cost about $0.012 per one million change records for databases hosted in the company’s U.S. regions, with slightly higher prices in Europe and Asia.

Apple pulls the plug on its €850M data center project in Ireland over planning delays

Dark clouds have gathered and broken over Apple’s plans to build a data center in Ireland. Three years ago, Apple announced that it would invest $2 billion into building a pair of new, green data centers in Ireland and Denmark. But today, the iPhone giant confirmed that it was cancelling the first of those two projects, after too many delays in the approval process, which today appeared to be extending in a way that could go on for a long time to come.

“We’ve been operating in Ireland since 1980 and we’re proud of the many contributions we make to the economy and job creation.  In the last two years we’ve spent over €550 million with local companies and, all told, our investment and innovation supports more than 25,000 jobs up and down the country.  We’re deeply committed to our employees and customers in Ireland and are expanding our operations in Cork, with a new facility for our talented team there,” the company said in a statement provided to TechCrunch. “Several years ago we applied to build a data centre at Athenry. Despite our best efforts, delays in the approval process have forced us to make other plans and we will not be able to move forward with the data centre. While disappointing, this setback will not dampen our enthusiasm for future projects in Ireland as our business continues to grow.”

Apple had planned for the data center — which would cover 166,000 square metres — to go online in 2017.

(The first phase of the Danish center announced at the same time, incidentally, is nearly completed and Apple is now working on a second center in the country. We’ve confirmed with sources that this second center is not the “other plan” that Apple refers to in its statement above, meaning another data center announcement from Apple in the region may be coming.)

As originally conceived, the facility in Ireland was planned to be built on land previously used for growing and harvesting non-native trees. As part of its CSR in building the facility on that land, Apple also pledged to “restore native trees to Derrydonnell Forest,” as well as build an outdoor education space an a walking trail.

But within months of Apple announcing the project, issues started to arise around the potential environmental impact and what effect the building of the data center would have on the national electric grid. Initially, the Galway County Council asked for more details from Apple about how the data center would work.

Then, when Apple provided it and the council granted permission to build the center six months later, individual objections started to surface, including from a local environmental engineer called Allan Daly, who has become something of the public face of the opposition to the plans.

Daly’s main argument was that Apple’s data center, particularly at its fullest-possible size, would put too much strain on Ireland’s power grid, including in the building of them. Apple has maintained that its data centers are powered on renewables, that it gives back over time, and that it wouldn’t over-build.

Last October, Apple won a case in the Irish High Court that appeared to give the company the green light it needed to proceed with its plans. But from what we understand, there was still some uncertainty that lingered, because opposition could have still taken the case to the Supreme Court to appeal once again.

That continued uncertainty was the final straw for Apple. With no guaranteed end in sight, Apple finally made the choice to “move on”, as one source close to the situation told TechCrunch.

The whole case underscores some of the ongoing issues that apparently exist in Ireland over how data centers are planned and approved by local authorities.

“There is no disputing that Apple’s decision is very disappointing, particularly for Athenry and the West of Ireland,” Ireland’s Minister for Business and Enterprise Heather Humphreys said in a statement provided to Reuters.

There is talk of reforming that whole process, but that is not something Apple will get involved with at this point.

The company has had a rather complex relationship with the country.

Like many tech companies, Apple has made a lot of investment into operations based out of Ireland, including housing its European headquarters in Cork. But the country has also been the subject of a large tax debate, which has seen Apple just weeks ago finally settle on paying some €13 billion ($15.4 billion) in back taxes to Ireland starting this month, after the EU ruled that the existing tax scheme was illegal.

Ironically, Ireland was on Apple’s side in trying to resist the payment — perhaps in part because it all too well understands its relationship to the companies that subsequently pump hundreds of millions of euros in investment and jobs into their economies.

It’s odd timing, therefore, that we’d hear about Apple pulling out of the data center in Ireland now, although from what I’ve been told the two are very distinct, unrelated issues.

Necto looks to help individuals get their own local ISP businesses off the ground

If you live in a city, you’re probably deciding between a handful of major broadband or wireless carriers — maybe something like Comcast or AT&T. But there’s a good chance that there are a bunch of local carriers that are looking to get off the ground, and Benjamin Huang wants to help make sure there are even more options/.

That’s the idea behind Necto, a startup looking to create a sort of ISP school to help people get started with their own internet service provider founded by Huang and Adam Montgomery. Typically that’s a pretty tall order, but Necto works with individuals to learn how to build a network, get the right equipment, and deploy it in order to get consumers access to a new internet service provider that’s an alternative to the larger carriers. There are already emerging providers like Sonic in San Francisco, which aims to offer quick internet for a cheaper price, but there’s a whole group of individuals waiting in the wings that are trying to build their own ISP and the associated business behind it, Huang said. Necto is launching out of Y Combinator’s winter 2018 class.

“Ultimately, we want to see so many ISPs that net neutrality isn’t an issue,” Montgomery said. “It’s cheaper than ever and easier to start an internet service provider. People didn’t know they could do this, and networking engineering is the highest cost. You have to have a lot of stuff to build out. We remove that and bundle it as an ISP starter kit service. We give guidance to the operators, these are the customers you have, this is the equipment you need buy, here’s how to construct them. It’s more like constructing Ikea furniture. The hard part we remove which is automatically configuring these routers.”

Necto started off as its own attempt at an internet service provider, but Huang and Montgomery found that trying to get wholesale fiber was a high barrier to entry. The pair started looking into wholesale wireless, and Huang said that technology is getting to the point where it’s just as fast as typical broadband and an option for resale. The challenge then is getting the equipment into the hands of individuals that want to ramp up their own ISP and showing them how to get started. Then, they’re off to the races and work to build a business around that, including customer service and other facets of it.

Necto essentially charges for the guidance of how to start an ISP, including a class that individuals go through in order to get one off the ground. Then the company continues to ship software to ensure that it’s not as difficult to keep the equipment up and running, as well as provide ongoing support for those individuals. The equipment is all off the shelf, Huang said, in order to lower the barrier to entry for these providers.

The challenge here, however, will be ensuring that not only individuals know they can get an ISP off the ground, but getting their — and consumers’ — attention in the first place. Necto hopes to take a hyper-local strategy, Montgomery said, like traveling to farmers’ markets and working with local operators to ensure they can track down the right people that are looking to build a business around ISPs. There are still going to be plenty of challenges as it continues to work with wholesale wireless providers in order to get these businesses off the ground.

Duplex shows Google failing at ethical and creative AI design

Google CEO Sundar Pichai milked the woos from a clappy, home-turf developer crowd at its I/O conference in Mountain View this week with a demo of an in-the-works voice assistant feature that will enable the AI to make telephone calls on behalf of its human owner.

The so-called ‘Duplex’ feature of the Google Assistant was shown calling a hair salon to book a woman’s hair cut, and ringing a restaurant to try to book a table — only 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 comfortable“, as Google couches it in a blog about its intentions 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 end of the line that they were speaking to people.

At one point the bot’s ‘mm-hmm’ response even drew appreciative laughs from a techie audience that clearly felt in on the ‘joke’.

But while the home crowd cheered enthusiastically at how capable Google had seemingly made its prototype robot caller — with Pichai going on to sketch a grand vision of the AI saving people and businesses time — 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 experiments do appear to have been designed to deceive,” agreed Dr Thomas King, a researcher at the Oxford Internet Institute’s Digital Ethics 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 deception and not the user experience… You don’t necessarily need to deceive someone to give them a better user experience by sounding 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 intention is — but… even if they don’t intend it to deceive you can say they’ve been negligent in not making 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 a university 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, humans 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 put 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 conversations troubled the I/O stage, however.

Yet Pichai said Google had been working on the Duplex technology for “many years”, and went so far as to claim the AI can “understand the nuances of conversation” — albeit still evidently 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 understanding, 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 phone calls, but Pichai described the calls as “real” — suggesting 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 examples where the calls quite don’t go as expected but our assistant understands the context, the nuance… and handled 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 — a feature 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 well understood competition. Nor was there a set of rules that everyone was shown and agreed to beforehand (at least so far as we know — 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 jobs. Can you see 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 say look 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 preferred 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 say well consumers prefer this slightly deeper shade of blue to a lighter shade of blue.”

Had Duplex been deployed within Turing Test conditions, King also makes the point that it’s rather less likely it would have taken in so many people — because, well, those slightly jarringly timed 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 immediately 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 only represent a small minority of the population speaking in ‘BBC English’ and so, perhaps in a sense, using 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 risks, 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 time 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, maybe, as you normally would, and that could potentially 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 way 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 potential applications. We’re really just getting started.

But if the world’s biggest and most powerful AI developers believe it’s totally fine to put ethics on the backburner then risks are going to spiral 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 risk 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 put it, experimental — use of Duplex will be to supplement Google’s search services by filling in information about businesses’ opening times 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 humans (and their associated businesses and institutions) 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 spokesperson. 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 said Google wants to “work hard to get this right, get the user experience and the expectation 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 break it, later fix it’ ethos of some of the tech industry’s major players…

Google Assistant making calls pretending to be human not only without disclosing that it’s a bot, but adding “ummm” and “aaah” to deceive the human on the other end with the room cheering it… horrifying. Silicon Valley is ethically lost, rudderless and has not learned a thing.

— zeynep tufekci (@zeynep) May 9, 2018

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 point where even politicians 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, deception 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 out a first draft of a framework to guide ethically designed AI systems at the back end of 2016 — which included general 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 standards body developed a specific standard — BS 8611 Ethics design and application robots — which explicitly names identity deception (intentional or unintentional) as a societal risk, and warns that such an approach will eventually erode trust in the technology. 

“Avoid deception due to the behaviour and/or appearance of the robot and ensure transparency of robotic nature,” the BSI’s standard advises.

It also warns against anthropomorphization due to the associated risk of misinterpretation — so Duplex’s ums and ahs don’t just suck because they’re fake but because they are misleading and so deceptive, and also therefore carry the knock-on risk 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 “only for well-defined, limited and socially-accepted purposes”. (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 need 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 achieve: It could just have programmed the assistant to say 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 showing Duplex pulling the wool over busy and trusting humans’ eyes — and by doing so showed 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 use 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 raises 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 subject 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 deception, robot addiction and the potential for a learning system to exceed 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 shown the same tired tech industry playbook applauding engineering capabilities in a shiny bubble, stripped of human context and societal consideration, and dangled in front of an uncritical audience to see how loud they’ll cheer.

Leaving important questions — over the ethics of Google’s AI experiments 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 sorts 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’ paradox may explain Google’s choice of deception for its Duplex demo: Humans don’t necessarily 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 reveals itself to be a robot the human who picked up the phone might well just put it straight back down again.

“Going back to the deception, 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 necessarily use it in a context where people really 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 makes much sense.

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

So really 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 cost 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 makes the point that the Duplex system is (at least for now) computationally costly. “Which means that Google cannot and should not just 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 contexts — and they can also guarantee it will only be used with certain safeguards built in. So I think the experiments are maybe not the best of signs but the real test will be how they release it — and will they build the safeguards that people demand into the software,” he adds.

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

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

It really makes you wonder whether, at some foundational level, Google lacks trust in both what AI technology can do and in its own creative abilities to breath new life into these emergent synthetic experiences.

Apple invests $10M in greenhouse gas-free aluminum smelting

Canadian Prime Minister Justin Trudeau and Quebec Premier Philippe Couillard joined key execs from Apple and industrial manufacturers Alcoa and Rio Tinto to announce a new process for smelting aluminum that removes greenhouse gases from the equation.

Alcoa and Rio Tinto are creating a joint venture in based in Montreal called Elysis, to help mainstream the process, with plans to make it commercially available by 2024. Along with swapping carbon for oxygen as a byproduct of the production process, the technology is also expected to reduce costs by around 15 percent.

It’s easy to see why Apple jumped at investing into tech here, pumping $13 million CAD ($10 million USD) into the venture. The company has been making a big push over the past couple of years to reduce its carbon footprint across the board. This time last month, Apple announced that it had moved to 100-percent clean energy for its global facilitates.

“Apple is committed to advancing technologies that are good for the planet and help protect it for generations to come,” Tim Cook said in a release tied to today’s news. “We are proud to be part of this ambitious new project, and look forward to one day being able to use aluminum produced without direct greenhouse gas emissions in the manufacturing of our products.”

Those companies, along with the Governments of Canada and Quebec have combined to invest a full $188 million CAD in the forward looking tech. While the new business will be headquartered in Montreal, U.S. manufacturing will also get a piece of the pie. Alcoa has been smelting metal through the process at a smaller scale in a plant outside of Pittsburgh since 2009.

House Democrats release more than 3,500 Russian Facebook ads

Democrats from the House Intelligence Committee have released thousands of ads that were run on Facebook by the Russia-based Internet Research Agency.

The Democrats said they’ve released a total of 3,519 ads today from 2015, 2016 and 2017. This doesn’t include 80,000 pieces of organic content shared on Facebook by the IRA, which the Democrats plan to release later.

What remains unclear is the impact that these ads actually had on public opinion, but the Democrats note that they were seen by more than 11.4 million Americans.

You can find all the ads here, though it’ll take some time just to download them. As has been noted about earlier (smaller) releases of IRA ads, they aren’t all nakedly pro-Trump, but instead express a dizzying array of opinions and arguments, targeted at a wide range of users.

“Russia sought to weaponize social media to drive a wedge between Americans, and in an attempt to sway the 2016 election,” tweeted Adam Schiff, who is the Democrats’ ranking member on the House Intelligence Committee. “They created fake accounts, pages and communities to push divisive online content and videos, and to mobilize real Americans,”

Russia sought to divide us by our race, our country of origin, our religion, and our politics. They attempted to hijack legitimate events meant to do good – teaching self-defense, providing legal aid – as well as those events meant to widen a rift.

Here’s just some examples: pic.twitter.com/YMX2FTgPGU

— Adam Schiff (@RepAdamSchiff) May 10, 2018

He added, “By exposing these Russian-created Facebook advertisements, we hope to better protect legitimate political expression and safeguard Americans from having the information they seek polluted by foreign adversaries. Sunlight is always the best disinfectant.

In conjunction with this release, Facebook published a post acknowledging that it was “too slow to spot this type of information operations interference” in the 2016 election, and outlining the steps (like creating a public database of political ads) that it’s taking to prevent this in the future.

“This will never be a solved problem because we’re up against determined, creative and well-funded adversaries,” Facebook said. “But we are making steady progress.”