BY STEVEN LEVY
It is a gorgeous late summer afternoon, and I am sitting with Ray Ozzie in his spacious home office in Manchester-by-the-Sea, 30 miles up the coast from Boston. The software visionary who created Lotus Notes and who later succeeded Bill Gates as Microsoft’s chief software architect, is explaining to me how the humble phone call is not dying, as many might believe, but is busy being reborn.
It’s not an abstract subject for the 58-year entrepreneur. For the past few weeks I have been using the app his company is announcing today, called Talko. It’s a weird, almost magical, combination of phone calling, text messaging, virtual conferencing and Instagram-ish photo sharing. Depending on how you view it, Talko is three or 39 years in the making.
At one point, Ozzie wants to show me something on the app. We both pull out our iPhones and connect with each other; actually, in that moment, we reconnect to a conversation we’ve been having all month that’s been recorded and archived in the app. I think my editor might be interested in the discussion, so we expand the conversation to include him. He’s unable to join us at the moment—I should have known, because the app lets me see that he’s walking around somewhere on the West Coast—but I shoot a photo for him to look at anyway, and Ozzie and I continue talking. Later, my editor will listen to that part of conversation and see the picture at the moment we shot it. And he’ll have the option to comment, perhaps kicking off a longer discussion down the road, either by convening us together in real time or continuing in the same piecemeal fashion as today.
That phone call represented an amazing advance in communications. But Ozzie considers it equally amazing that in the 139 years since “Mr. Watson, come here,” phone calls haven’t changed much. In an era of dazzlingly accelerated communications, we are still stuck in the voice paradigm of a live conversation initiated by one person in hopes of dominating the attention of a recipient. (Telecom insiders have a name for this: “caller hegemony.”) No wonder we still cling to obsolete jargon like “dial,” “ring,” and “pick up the phone” when none of this exists in the actual technology any more. Meanwhile, mobile users increasingly eschew telephony for more flexible formats like email, texts, and photo apps; traditional phone calls themselves are somewhat of an endangered species. If you’re under 30, the very idea of getting a cold-call from someone besides a close friend or family member seems intrusive, if not borderline odious. Pundits are talking about the “death of voice.”
“It’s not like voice is an unimportant data type,” he says. “We have mouths! We have ears!” What’s more, he adds, we carry around incredibly advanced tools equipped with fantastic microphones and advanced speakers. Yet many of us have abandoned the ancient practice of speaking to each other for email and text, where nuance is often lost and bullshit too easily rules. On a whiteboard, Ozzie has jotted down a long checklist of emotions easily conveyed by voice, but difficult to decipher in quickly thumbed-out bursts of text: concern, pain, urgency, empathy, clarity, seriousness, confidence, anxiety, trust, strength, accountability, anger, fear, stress, confusion, doubt…
Gaps like this mean opportunity, and Ozzie’s solution is Talko, his ploy to restore the human voice to primacy in the digital age. The app’s key innovations allow people to build conversations as if they were collaborative documents — adding to these artifacts both in real time and asynchronously. (The most striking aspect of Talko is how seamlessly it mixes live interactions and monologues into a single stream.) And in another paradigm shift, everything is archived by default, allowing people the same persistent access to their meetings and conversations that they now have with email and documents.
Basically, Ozzie has tried to build telephony as if it was being invented from scratch, assuming the Internet and mobile phones as basic building blocks. His focus, in keeping with a life-long devotion to using communication to facilitate productivity, is providing tools to make workgroups more effective. But he believes that the innovations he’s unveiling in his product —which launches today on iPhone — can enrich everyone’s life.
How can an app displace 135 years of telephony?
The first step is by being an app and not a facsimile of a desk phone. Unlike the “dialer” on your mobile, or even the call button on your contact list, Talko presents the opportunity to begin a conversation by simply touching on the image of the person or team you want to talk to, whether they are ready to join right now or sometime in the future.
Even before you make the connection, Talko gives you some information that your Princess phone or even your iPhone never provided: Using the sensors on a recipient’s phone, icons tell you whether that person is on the move (walking or driving), how the person is connected to the network, and even what time zone she’s in. It’s a pretty good way to gauge receptivity to a live call.
But no worries either way, because Talko makes it easy to jump from live communication to thought. If you want to avoid the awkwardness of a cold call, you could simply start by texting a message, or just talk into the microphone without “ringing” the recipient. Whether you’re communicating in real time, or adding to a conversation on your own, you can instantly add a photo to the stream (in later versions Talko will let you add video and documents). All that activity is later available simply by checking the log of the call — you can dive in at any point in the conversation and find what interests you, especially if you or someone else has pinpointed a crucial moment, by leaving a bookmark.
“What’s striking is the fluid nature of going from synchronous [real time, like a phone call] to asynchronous [discrete, like text or email],” says former Netscape co-founder Marc Andreessen, whose VC firm has invested in Talko. “Synchronous communication is great but there are times when you want to drop into voice, or text. Talko is a big step forward in finding that balance.”
That fluidity turns caller hegemony into something more like an improvisational jazz combo. If you are part of a call that “belongs” to a group of four people, you might turn on your phone and discover that a live conservation is occurring among two or three of them. You’d have the choice of joining in or waiting to catch up later—the communications equivalent of binge-watching a television series after it’s finished its original run.
Talko is primarily designed to work with groups, in particular to boost the productivity of collaborative workplace teams. The Talko engineeers, who have been testing the product for over two years, use it for all of their stand ups, the regular meetings when they discuss progress, flag problems, and set short-time tasks. It’s not only invaluable to those who might miss a session, but also to people who are in the room together for the meeting, who can refer to key moments in a playback.
But Talko is also geared for wide use among the general population. “We will only succeed if it is usable by everybody. It can’t feel like a business tool while you’re using it,” says Ozzie. He and his colleagues have been testing it with their families for over a year and it has totally replaced traditional calling for them. “We talk way more, infinitely more than we ever used to talk with each other,” he says.
Some futurists, bucking the “death of voice” meme, have been long anticipating a renaissance of voice via Internet. One of them is Martin Geddes, a consultant who co-founded the Hypervoice Consortium, an industry group funded by various Internet voice concerns. Based on a look earlier this month, Geddes says Talko is the best voice-based product he’s ever seen. “I’ve been saying for a long time the revolution is coming in computer-assisted conversation,” he says. “Things like Talko will turn us into lords of time and space.”
Talko is built so anyone can easily grasp its operations. But to understand its origins, and why this company has a chance to finally drag telephony from the Bell and Watson days, you first must know about Ray Ozzie. While he’s not a household name, he is one of the towering figures of the software world. And throughout his career, he has manically focused on one area.
Ozzie first became smitten with this vision at the University of Illinois when, as a sophomore in 1975, he stumbled upon an innovative mainframe computer system called Plato. Created to promote “automated learning,” the inventors had devised a suite of tools that foreshadowed many of the ways that people would use digital technology to communicate and collaborate decades later – electronic mail, texts, games, and even a voice chat program called Talkomatic. (Does that sound like the name of this new app? No coincidence.) It also allowed programmers to write code interactively, a welcome change from the punch-card regime Ozzie had been used to. Ozzie spent much of the rest of his college career in the artificial Nirvana of the Plato community, where hundreds of people indulged in the future via the system’s iconic orange terminals.
After graduation Ozzie built applications for companies pioneering software for the then-emerging personal computer. (Some people still fondly remember Symphony, a suite of apps written for Lotus Development Corp., then an industry leader.) But his dream was to replicate the collaborative magic of Plato on the personal computer, which had only recently been invented. His vehicle was Lotus Notes, a massively ambitious “groupware” system released in 1989 after five years in development. It became the company’s flagship product. Indeed, Notes was the prize that led IBM to buy Lotus for $3.5 billion in 1995. Ozzie led the Notes division at IBM for two years, helping the product grow to what would eventually become a base of over 150 million. (The product still lives, as “IBM Notes.”)
Ozzie is happiest as a builder, and he left IBM to create yet another version of Plato Heaven, this time updated to reflect Internet ubiquity. He gathered a small team on the third floor of the mansion, and they began Groove, envisioned as a virtual gathering space where people could participate live or piecemeal, sharing documents and managing projects. Introduced in 2000, Groove was just getting traction when the tech bubble popped. The inevitable buyer was Microsoft, in 2005.
Bill Gates had long been an Ozzie admirer, once calling him “one of the five best programmers in the universe.” He made no secret that bringing Ozzie to Seattle was as much his motivation as buying Groove’s technology. Gates was stepping back from Microsoft’s operations, and there was a sense that the visionary torch was being passed. One indication of that came when Ozzie wrote a come-to-Jesus memo called, “The Internet Services Disruption,” basically a wake-up call to Microsoft that the cloud era was coming and the company better get skyward. It came almost exactly a decade after Bill Gates circulated a similar memo heralding the Internet age. Gates himself hit the send button to distribute Ozzie’s memo company-wide. Eventually Gates relinquished his title of Chief Software Architect —and all the attendant baggage of succession—to Ozzie.
But Ozzie’s approach of exercising “soft power” had limited success in Microsoft’s highly political, top-down culture. Ozzie devoutly believed that great software is built by small groups, but it was a Sisyphean task getting the corporate behemoth to budge from its human-wave approach to development. He got pushback when he tried to implement his philosophy, even when trying to re-engineer workspaces to make them more collaborative.
He did manage to build the company’s cloud strategy, mainly by taking a tight cadre of engineers working outside the server division to build a new system. At a certain point the outsiders became the server division. But he had very little luck in his attempts to revive Microsoft’s consumer division, as the company missed the mark on music players, phones and tablets.
“I was very proud to move the needle three degrees, in the sense of moving the course of a supertanker,” he now says. But as the five-year anniversary of his memo approached in 2010, he began negotiating an amicable exit with CEO Steve Ballmer. “I just wanted to go out and build again,” he says. He was gone by the end of the year.
But what would he build? He spent most of January and February 2011 in the Bay Area, hanging out in the offices of former Lotus CEO Mitch Kapor’s software foundation. He very consciously sought out some of the best minds of the Internet startup boom, like Dropbox’s Drew Houston and Medium founder Ev Williams. (Disclosure: you are reading this on Medium!)
In truth he was already zeroing in on his next project, based on a core idea that it was time for the human voice to reassert itself online. One contributing factor to that idea might have been what he saw at Microsoft: a communications breakdown rooted in part by an overreliance on email.
“The email culture runs deep at Microsoft,” says Ozzie. “Email is an amazing thing. It enables communication, but allows you to posture in a way you would never do face to face.” Hiding behind the conventions of email, it’s possible to passive-aggressively ignore a problem, or subtly orchestrate a position. All too often, even small groups are able to skate around disputes they would have hashed out had they been forced to interact in real time. “Things can get dysfunctional pretty quickly when people don’t talk to one another,” he says. “Microsoft had a corporate phone system, but no one used the phone — people used email to build this wall of text. If it had been a phone call culture from a different era, it would have been a totally different dynamic.”
That led Ozzie to begin thinking of how technologies like email and text had diverted all of us from the rich experience of talking to each other. Ultimately, he asked himself the question: “What if the phone had never been invented and the Internet came first? And then someone said, ‘Hey, we can put voice on this thing!’”
In April of that year, Ozzie joined a cadre of tech experts providing technical aid in the wake of the tsunami that hit Japan. He spent a lot of time in Asia, largely helping to set up a crowd-sourced network of radiation reporters called Safe-Cast. But he kept thinking about the voice idea, developing it in his head. And on April 19, 2011, flying home from Tokyo, he sketched what his app would be like to users. Not long after, he filed for some patents.
Toward the end of the year, some of his former collaborators pinged him to see if he was working on something new. One, engineer Matt Pope, had earlier followed Ozzie from Groove to Microsoft. Another, Eric Patey, dated even earlier, to Ozzie’s original company (sold to Lotus in 1994), Iris Associates. Both became Talko co-founders. Over the next few months, Ozzie brought in some other refugees from his previous projects. “We got the band back together,” he says.
The team reunited near Ozzie’s home in Manchester-by-the-Sea, gathering at a rented office in the second floor of a 129-year old fire station. They also had some strategy meetings in Ozzie’s office on the third floor of a seaside Gilded Age “cottage,” which in the intervening years had become kind of a mini-computer history museum. Thus some of Talko was brainstormed by a small team surrounded with antique devices ranging from a 1950s steampunky Edison “Ediphone” dictation machine (once used by Ozzie’s father) to a 1990s General Magic Communicator. And of course, a working PLATO terminal. Ozzie insisted that the company, though founded by adults of a certain age, should build its product on the model set by modern startups led by peachfuzz founders: open-source tools, servers hosted by Amazon Web Services and Microsoft Azure, small teams of workers in different locations. “We talked a lot about how, to do it right, we would have to start from scratch, both personally and operationally,” says Pope, who spoke to me via Talko. (It’s pretty good way to conduct an interview.) “The first six months we were learning all we could about open source packages and how distributed teams communicate.” (The Talko team is geographically scattered, with workers in Seattle, Boston, San Francisco, and Manchester-by-the-Sea.)
Most of 2012 was spent tackling the technical issues (one aid: a new open source technology called WebRTC, which makes writing reliable voice-based Internet software much easier), and by 2013, they began to shift the focus onto user experience, eventually building to about 250 testers, including around a dozen business teams. Drawing from the feedback, they iterated constantly. Earlier this year, they made a major change that shifted the focus on the home screen from a list of conversations to the pictures of potential contacts. “That was a magical inflection point,” says Pope. “It optimized you into getting into a conversation quickly.”
There’s still plenty to do: as of now, for instance, Talko only works with iPhone, and a product meant to connect everyone can’t really afford to ignore the world’s most popular mobile operating system. Video has yet to be implemented. But Ozzie wants to learn now from seeing Talko exposed to a mass audience.
While Talko offers a number of compelling features, one in particular is destined for controversy: it stores all conversations by default. Unless you are being bugged by the FBI, the spoken words of a phone call have always disappeared as soon as the sound subsided. In that sense, traditional telephony is more ephemeral than Snapchat. But Talko calls are persistently available, like email or texts.
At this point, one envisions privacy advocates gathering feathers and warming up the tar. But Ozzie and his team have thought a lot about how privacy concerns play out. (Ozzie himself has a history of fighting for consumer protections and is a board member of the Electronic Privacy Information Center.) At any point, any participant in the conversation can delete the storage, vaporizing the conversation. Even long after the exchange is over, anyone in a Talko call retains the ability to erase a segment in which he or she has participated.
(In most cases, the point will be moot because many conversations will in effect disappear; even though everything on Talko is archived, when the system gets out of beta, storage will be temporary for most users—after a week or so, an icon will appear beside a conversation indicating that it is about to be recycled. If users want to retain the file, they can download the conversation. But they can also pay Talko to have unlimited storage, and one assumes that many will do so.)
Of course, those stored conversations might be available to government subpoenas, national security letters and perhaps even NSA voodoo. But the bet is that customers will understand that having a record of conversations is an invaluable thing. Walter Zimbeck who works for a consulting operation called Technology Assessment and Transfer, has been beta-testing Talko. He says he’s constantly accessing the archived conversations, even going over the voice messages he’s sent so that he can improve his own communication. Another tester says that group meetings are much less stressful because he knows he’ll have a chance to play back the session to see what he missed. To make this process easier, some Talko beta testers have begun using topic hashtags using keywords used during conversation, so participants can easily search for that moment later on.
Those using Talko outside the business context may find the storage aspect even more transformative. “Every call I made with my wife, my son, my daughter is in Talko. Every call with Matt Pope, every daily stand up,” says Ozzie. “I could play for you the most amazing calls, where my daughter announced her pregnancy and my son announced he was engaged. Have I deleted some heated calls that shouldn’t be kept around? Absolutely. But I’ve never been looking for a call and not found it.”
Geddes of the Hypervoice Consortium is impressed at how seriously the Talko team has considered the privacy issue. “Giving participants the ability to rescind is critical,” he says. In fact, Geddes sees Talko, especially in the way it balances real-time and asynchronous participation as an antidote to Caller Hegemony. “The power is rebalanced,” he says.
Ozzie sees Talko making its money with premium services to workgroups and corporations. The company is currently funded by a small group of investors including Andreessen Horowitz, Reid Hoffman via Greylock Partners, Kapor Capital and others.
Andreessen, who considered Ozzie one of his heroes even before working with him on the HP board of directors, thinks Talko will be a very big deal. “It’s hard to figure out any way where it’s a small deal,” he says.
Of course, it’s not a given that Talko will be the instrument of voice’s revival. Others have been down that path, without moving the needle even three degrees. For instance, a couple of years ago a well-backed startup called Sidecar introduced an app that promised to…reinvent the phone call. It didn’t take off, but co-founder Rob Glaser (better known as CEO of Real.com) says, “I still believe there’s an open opportunity.” Glaser thinks the key will be reliability and a network effect that will kick in only when many millions of people start using it.
Talko can’t hope for ubiquity until it gets an Android version and a full-featured Web version. (The corporate beta testers I spoke with sighed wistfully when reporting that a couple of team members could not participate because they didn’t use iPhones.) “We can’t build a business on this until we have all of the above,” says Ozzie, who wanted to make sure his engineers aced the iPhone first; now the team is well along on native versions on the other platforms. Even then, an independent app like Talko will have a tougher time winning a huge audience than a similar effort launched by Google or Apple, who could build it more deeply into the phone. (In last week’s iOS 8 rollout, Apple gave iMessage users the capability to send quickly dictated audio bites as texts—a very Talko-like feature.) But Ozzie feels that Talko’s benefits—in the workplace and beyond — cannot help but go viral. He points to the success of WhatsApp and Snapchat as proof that independent products from small companies can reach huge audiences. “My aspiration is that everybody who’s got a phone will use this tool or something like it to talk to one another,” he says. “There’s no reason not to.”
No matter what happens to Talko, it is remarkable that a pioneer from the floppy disk days has created one of the year’s most innovative mobile apps. Back in 2000, Bill Gates noted that, of the early visionaries of personal computing, very few were still in the game. “Except for Jobs and myself and Ray Ozzie, it’s not very many people,” he told me. Fourteen years later, Jobs is dead and Gates is no longer a full-timer at Microsoft.