How has the recent turmoil within the OpenAI offices changed your plans to use GPT in a business process or product in 2024?
Increased uncertainty means we are more likely to evaluate alternative AI chatbots and LLMs.
No change in plans, though we will keep an eye on the situation.
With Sam Altman back in charge, we are more likely to go all-in with GPT and LLMs.
What recent turmoil?

The New Stack Makers: Marc Canter with Memories of Scoble, Winer and The Internet of Things

Nov 18th, 2014 6:00am by
Featued image for: The New Stack Makers: Marc Canter with Memories of Scoble, Winer and The Internet of Things

In part one, Marc Canter makes his way through the meat shops of Chicago politics and learns to eat ribs. He gets to dance to the Talking Heads at CBGB in New York, moves to San Francisco and becomes one of the first to join the Apple developer community. In part two, Marc is still in San Francisco and making art on Potrero Hill. Canter is now the founder of Interface, Inc., an app authoring platform designed to take advantage of contextual data coming off of devices and sensors.

Marc Canter Interview – The New Stack Interview, Part Two

Marc: Throughout the rest of the 1990s, I ignored the web and I focused on what I saw was the evolution of multimedia: interactive TV. I was doing closed-circuit things. I built, on my home on Potrero Hill, a development grid, and I created an operating system…

Alex Williams: Where were you on Potrero?

Marc: 19th and Rhode Island. Triple lot. Total party home. We built the thing called the MediaBar – it was an operating system for location-based entertainment. You could do food and drink ordering. You had a music jukebox, a video jukebox. You logged into the thing, it grabbed your face. You had a drag-and-drop desktop. No one can ever patent human face in software because I had that working in ‘96. So I’m doing all this development work all on my own dime because I just left this nightmare scenario of these VCs and these people telling me what to do. So, fuck that – I’m going to do what I want to do. I got money.

So what happened to the company?

Marc: Which one? That company? That product? MediaBar?

Take me through again what happened to Macromedia.

Marc: So, it was Macromind; they didn’t have enough revenues; they merged with Paracomp, so it became Macromind-Paracomp. And this is all being done by a guy named Tim Mott. He was the hatchet man that came in to get rid of (John) Scull. We merged that and then he goes, “I don’t want to be CEO,” so he merged again with a company called Authorware. That became Macromedia.

And how do I know they were going to get rid of me? We moved from 5th and Townsend to 7th and Townsend. I go into the new office; there’s no desk for me and there’s no parking place. “Oh, did we forget to tell you that you’re fired? Oh sorry.” That’s the kind of people these were. By the way, Macromedia at around ’98-’99 had a 50% turnover rate. It was one of the most reviled companies in the industry. Anyway, so that’s Macromedia.

I took my money; I tried to retire for three months. I couldn’t do that. Fujitsu had offered me to become the president of Fujitsu America and the fucking board told me not to do it, so I listened to them. I was minister without portfolio. What I wanted to do was to invent interactive music videos. I hired Todd Rundgren, a guy named John Sanborn, and I spent a ton of money building the “Mediaband” – we had 500 auditions for this Todd Rundgren cyber band.

That was a big time.

Marc: Yes. Then there was ION, which did work with Bowie and Brian Eno. There was a Prince disc, a Bob Dylan disc; there were all these discs. We did a live version of the Mediaband that had a rack of MIDI – 96 channels of MIDI, and four break-out boxes that you could do constrained filtering on – switching between the tracks, and such that it will only switch on the beat.

The Mediaband led to The Marc Canter Show. Really, I wanted the show to be called [Insert Your Name Here] Show. Because I wanted it to be a template, so whether it was Chris Pirillo or Leo Laporte or Robert Scoble, I wanted a tool that allowed somebody to do iHazCupquake or all the Twitch stuff – this personality stuff. I was doing that fifteen years before the web.

The Marc Canter Show, then. And there was no way I was going to have the bandwidth, so I got two Sun machines, put them into my basement, and put a 100BASE-T network and 25 nodes in my house, and I got my own damn broadband network. I bought twenty-five Apple 8500s – state of the art. Each one had a camera on it and a laserdisc, because we didn’t really have digital video. I could do some digital video and be small, and if I wanted full-screen video, I triggered the laserdisc and it would play this stuff.

I hired caterers to do these dishes, so as you’d order food and drinks you’d see the dish being prepared. Nobody does that even today. I want to see how they’re working in the kitchen and what they’re doing.

So, fast-forward to today. You started a company.

Marc: In the late ‘90s I’m sitting with Dave Winer and Netscape calls and they invent RSS. He’s doing blogging; I’m an early blogger; and I came up with this notion of the “Community Commons” – it was an early notion of social networking. I took that to this guy Mark Pincus. He decides, “That’s great, but I just put money into Friendster. So let’s start something called tribe.” So it’s me, Pincus and Paul Martino. We started this thing called tribe. Oh – and danah boyd was there.

Was that the early days of FOAF?

Marc: Yes. And Paul Martino built FOAF into tribe. It’s awesome – I travelled to Galway, Ireland to join into the RDF stuff, Semantic Web, all this crap. I’ll spare you all the politics.

How did you meet Dave Winer?

Marc: I met Dave Winer at the Dallas MacWorld in 1986, and you have to imagine the scene: there’s this rodeo, and in comes – riding sidesaddle – Heidi Roizen, who’s now the top VC partner at (DFJ). She ran a software company called T/Maker.

We go outside and smoke a joint and that’s how I met Dave Winer. We’ve been best friends ever since.

What was Dave doing there?

Marc: A company called Living Videotext. First, they had ThinkTank, then they had MORE. And because he’s in the valley, he could suck up to Jean-Louis Gassée. He got Apple to buy 10,000 copies of MORE and every Apple employee got a copy of MORE, which was this cool Macintosh software.

I’m back in Chicago; I’m struggling to stay alive; I’m slammin’ gigs and stuff and doing anything I can until I got the VC money. Dave sold his company to Symantec. He then developed in the ‘90s one of the first desktop servers in a company called Userland Frontier, and then he built a blogging tool with that. That’s right when Meg and Ev started Blogger and the Trotts started Moveable Type.

The Trotts are totally running things until they decide to change their open source license, and this 19-year-old kid in Houston goes, “Well, I’ve got some PHP code.” I talked to him, like, that week:

“Who are you, dude?” “My name’s Matt; I’m going to drop out of college; I’m going to move to Silicon Valley; I call it ‘Wordpress.'”

Scoble was working for Userland then.

Marc: Yes, and then from Userland, as he likes to say, he fired himself. And then he went to work for Microsoft. I’m friends with all these guys. I have a photograph: we’re at the Brothers Korean place at 6th and Geary, when the world found out that Scoble had left Microsoft. I think he went to go work for John Furrier’s company. Dude, whatever.

photo of the very moment that the world found out that Scoble (who was Microsoft's first blogger) had left Microsoft

photo of the very moment that the world found out that Robert Scoble
(who was Microsoft’s first blogger) had left Microsoft

At the time, you know where the parties were: Arrington’s house. The Dropbox guys are living there; the Box guys are living there; there’s Ron Conway; Scoble’s in the tent naked; it’s the party scene. I’m like, look at this; this whole thing was bizarre to me.

There’s this moment with tribe where Bono comes in; Bono had this thing called and he was running these things with Brad Pitt and George Clooney and “can’t we all live together and hug it out?” He goes to Pincus and he says, “I want to license tribe – I want to white-label license it.” And Pincus goes, “That’s great. And I want the tribe logo up on the top right next to” And Bono goes, “No. It will be at the bottom of the page below the fold and it will say ‘powered by tribe.’” Pincus told him to fuck off.

Bono was going to pay him $1 million. So, myself and two other people sitting in the room – we all went off and started white label social networking companies. I called mine “Broadband Mechanics.” I built 27 different networks. I did the US Army ROTC, Sacramento Kings, NVidia, Bell Canada – a long list of things – and they all failed.

They all failed? Why?

Marc: Because they all thought, “if we build it they will come.” See, what Zuckerberg did was he played this thing – he played scarcity; you couldn’t join Facebook.

By the way, at the same time, this guy named Sean Parker came in and he was pitching this thing called Wirehog, which was his next version of Napster. It’s kind of like BitTorrent and he wanted to embed it in the social networks, but he had to run it as a separate company in case he would get sued by Hollywood. That’s when I met Sean Parker.

So, Zuckerberg: you couldn’t join Facebook, which of course then meant that everybody wanted to join Facebook, and then he ingeniously opened it up. MySpace went through this meteoric thing, but we all knew what was going to happen once Murdoch bought MySpace. He was going to destroy it, which is exactly what he did.

Anyway, I’m building all these social networks; the final straw was when this ’07-’08 thing happened. I wrote this book; we’re doing all this activism at something called That happened to be released the same week as YouTube launched. So, there’s all these things going on, but all these networks were failing, and at that point I had been in the Valley for 21 years.

Two things happened; I was already pitching this other tool called a “Persona Editor,” but it was banked from the fact that we would have a two-way web, so I can get data out of the thing but I could also write the data back. The notion was: I’m going to bring this data in, I’m going to do something to it and send it right back to where it came from, and then it’s going to ricochet through me and I’m going to have processes and things. No one got it, no one wanted to fund it, and even today you couldn’t build it. I did a Google Tech Talk; if you type in “Persona Editor,” you’ll see this thing I did in ’09.

I picked up the family and left California, and I went to the least likely places you could imagine, because I wanted to get out of technology. I’ve been doing it for 21 years, and not including the time in Chicago and being a videogame guy. I mean, my entire life had been a technologist and this is echo chamber.

For five years I tried to do an economic development plan in Cleveland: it was partially education, partially marketing, and kind of the opposite of Groupon – hyper-local stuff. It was a citizen dashboard platform and an open source platform combined together. I took it to Jamaica, Tampa, Kansas City; Lafayette, Louisiana. Each place I went I progressed further and further, but I never got any funding for it.

I then started doing some work-for-hire stuff in ’13 in the EdTech bubble – all these new startups doing educational stuff. I get this gig being the acting CEO of an Internet of Things company. So, now I got first-hand knowledge; I’m going to meetings with Intel and Cisco; I’m listening to these guys.

Now, remember: I’m a sophisticated guy. I know the subtleties between why is Jolie O’Dell doing this, and someone did that and Arrington did this but he’s really doing that, and the power structure – Kara Swisher. Kara and Megan – we did the same Lamaze class together – I’ve known Megan since she was in school at MIT. So, I knew the scene; I could step back; I’ve seen the eras and the rollercoasters. I’m investigating this thing called IoT, and this is more than another bubble, because IoT – Internet of Things – you can’t have it without mobile and apps. You can’t have it without social, and you surely can’t have it without digital media. I mean, it’s everything.

So, that job lasted three weeks, because he was a nightmare – I won’t even mention his name.

So, I was the CEO of this company, and I came with this whole strategy, and I pitched it to Paul Martino. Paul was in on Zynga; he’s Bullpen Capital. Paul says, “Marc, that’s the Macromind of the Internet of Things.” So, I came up with the name “ThingFace” because we’re putting a face on the Internet of Things, and that lasted until last Friday.

This version I want to show you is Version Twelve. I’ve been iterating, tweaking, pivoting and trying to get it right. We’re going to designers. Okay, I think we got it right. I mention this to Josh Elman. He goes, “Dude, I’m trying to figure out what is ‘ThingFace.’ What does that mean?” I said, “That’s the good news, right? It’s so weird and quirky I get to define what it means.”

So, if you were going to compare those early days when you were living in Chicago and you started to use these tools to make new kinds of music, moved to California, you got funded and it was kind of larger multimedia – how do you compare that to what’s happening today?

Marc: Completely different scene. So, we joined this accelerator called Wearable World. These are the dudes that just did the Skully helmet. You sit in a room; there’s ten, twenty companies; they’re all talking to each other. These guys are going to be my pimps. They’re going to take me out there, dress me up, get funding. They learned from Arrington and Om Malik.

“Wow, the power structure – if I put on a conference I’ve got everybody coming there,” so they’re putting on conferences for IoT. They have the Wearable Wednesdays events, but that’s what they started with. Then they got into a blog and they actually made good money doing consulting. The Internet of Things is different than these prior eras – it’s consumer-facing and it’s pro-sumer but it’s also industrial. It’s also Hell; it’s these things that we used to never touch. There isn’t a big company out there that isn’t trying to figure out what to do. All the investors are coming to Wearable World. So, I came back going to one hub – one family – I can totally get all that stuff taken care of.

So, the connectors are just there.

Marc: Oh my God – I used to do it all by myself before. Because the industry was smaller, when I would climb down a stage there’d be a line of people handing me their card wanting to do business. At a party you’re discovering people, but it doesn’t seem to me that parties are as much about business and the brand and communicating with the product. “This company’s throwing a party.” “Great, where are the drinks?” If I’m going to throw a party, I’m going to make sure that everybody who comes to that event knows my product and what it’s about. We’re much more spunkier than entrepreneurial. Things like Lean or Agile: we were doing that, but we just didn’t have a name for it. We’d see the customer is modifying, tweaking – of course.

What hasn’t changed?

Marc: Customer is the most important thing. It’s the experience. This whole thing about the cliques and the on-boarding and all that: the product (should) do something that gives you benefit and value; I totally refute all this wasted time. So, in that sense, Google is really what is happening. Facebook is doing some connection; they’re not really doing anything on Facebook. Yeah, I’m marketing myself; that’s about it.

The other thing is: software companies would have multiple products; a product would have multiple features. Nowadays – and this started with Web 1.0 – everyone is racing towards their IPOs, and a company that’s one product – one feature – that’s it!

When the ‘00s came along – I call it “The 37signals School of Thought”: “We’ve gotta pare down, pare down, pare down – make it simple, simple, simple – it should only do one thing.” That’s called Basecamp. Have you ever been able to effectively run a project with Basecamp? No, it doesn’t have enough features in it.

So, that’s the tendency – these little apps, I call them “schmatties” – it’s a Yiddish word for a “little nothing.” You want to waste your time, that’s fine. I’m building tools; I’m a toolsmith. Someone who buys my tool can make money. Nowadays I can put up my site and you can use my tool for free. I couldn’t do that in the old days. I couldn’t go through a two-tier distribution system, have you walk into a store, and pick up the product and use it for free.

So, what do you want now compared to what you wanted then, now that you’re doing this?

Marc: Since we’re on the notion of principles, I took the idealism of my parents and my family. Software changes the world. I’m lucky enough to have changed the world. Now, I call it kind of like an Orson Welles complex, because I was young when that happened and I may never have another hit again, but I can go to Heaven knowing that I helped changed the world.

Nowadays, people are about making money: “Software’s the new Rock n’ Roll. I’m going to go become a rock star and make all this money,” and, “I’m famous because I made a lot of money.” Snapchat, Instagram – these products are not worth that much money, but if some VC wants to pay for it and Wall Street wants to game it, whatever. I’m about building products that change the world and help others make money. What a notion. That’s what I wake up with in the morning.

The technology infrastructure that you’re building today – what can you tell us about it?

Marc: We’re building an authoring tool environment that creates apps, so, it would be “App Authoring for Mere Mortals.” It’s great that there’s programmers – Code Academy, Udacity. By the way I’m good friends with Doug Rushkoff who wrote the book Program or Be Programmed. That’s awesome, but the majority of the people – I bet you there’s not one person in this restaurant who’s going to be a programmer.

If these people want to participate in Internet of Things – they want some crazy/wacky idea – our authoring tool environment will allow them to create contextually-aware apps – apps that know when something changes. The idea is that all these Internet of Things platforms or devices know how to do one or two things really well: turn on a lock, turn on a light, count some numbers as I move; that’s awesome. You can do all these nice, little, simple things. You can plug into IFTTT and you can have a trigger that triggers an action; that’s beautiful. But to me, that’s the beginning.

Along comes the schoolteacher, a trainer, a tour guide, a guy who installs A/V in your house. They’re not programmers, but you want to do something that connects these things together. How about a mom? You got all this gear in the house and the mom wants to make sure her kids understand this technology. I know my wife does. So, if the door opens, and the sprinkler thing, and you’ve done your homework, then you get the candy. I mean, you can imagine these elaborate Rube Goldberg things, fun things.

Do you know what percentage of Fitbits are set aside? It’s like 80%.

Now, that’s kind of unfair because those Fitbits are supposedly used for exercise and it’s not the Fitbit’s fault. It’s just the human doesn’t get motivated. So, to us, and because of the size of the fitness devices, no matter what, the first agent that we do will be a personal trainer agent. It will work with these Fitbits, and it will work with RunKeeper, or Runtastic, all these products are out there. They’re all doing what they want to do. We’re going to come up with new kinds of apps, new experiences that react to contextual changes that allow a personal trainer to encapsulate the essence of his being in the app.

The thing is: two humans interacting together, and the software’s in the middle. We’re not going to make software that’s supposed to stand on its own. We’re going to make software that needs a human to be part of it. So, I wrote a white paper: instead of it being Internet of Things it should be the “Internet of Humans.” I mean, we’ve totally missed that.

What’s the ecosystem, then?

Marc: The agents will be both public and private. A personal trainer does a one-off agent to his client. This trainer guy has five clients, and he’ll take the generic personal trainer thing and he’ll create instances of it. He’ll build one that will propagate content into it, and some rules. He’ll make it for people who want to gain weight, and he’ll do one for people who want to lose weight. He’ll do one for women who just had a baby and have extra baby weight, or geriatric…

The customer will go to the app store, get the generic app player. He’ll hand the client a code, and that customized app – that one-off app – will appear for that client. So, that’s a private agent. We’ll have a marketplace built into the tool. Let’s say one of our agents is a tour guide, in every region around, so many people in San Francisco…

People could share agents…

Marc: Absolutely. They can sell and buy and sell them. They’d have a marketplace…

You’re using “agents” as a metaphor now for…

Marc: …for the contextually-aware app.

As opposed to pages.

Marc: Right. That’s what Eightly does. Scoble demoed it with Andy Grignon at LeWeb back in December – these are the guys who worked on iPhone and the webOS – that’s still like a media channel, and that’s cool – it’s like a Flipboard thing, like an app that’s media, and it knows where the beacons are. Whatever – it takes all types, right?

As I was building these apps I realized: this is really personalized publishing. It’s really what Matt Mullenweg should be doing. It’s the next evolution of personal publishing, where I’m publishing just to you.

The agent can do two kinds of things. It can have something running in the background, be semi-autonomous smarts, but also can create front-end interfaces which is why we call the company “Interface, Inc.”

And those agents could have a small memory footprint.

Marc: Exactly, and so I could have one app that’s multi-tenant.

Very similar to what you’re seeing in program language development…

Marc: …But not a line of code will be needed. Completely visual programming.

But there are some similarities…

Marc: Absolutely, which is the way it should be. And the trick is coming with paradigms – authoring paradigms that humans can understand – and I think I’ve done that. If I want to create a foreground sequence, something that the user instigates, it’s a simple sequence of things. I can go, “ I want that to happen earlier,” or “what happens here?” Drag the icon. “Okay, it’s going to display this. It’s going to do this. It’s going to do a survey. It’s going to do a wizard.” That’s the foreground activities. Or, “I want a background thing. Pick up where IFTTT left off. If it’s raining, then move the running club to the ‘blah-blah-blah’ venue.”

How do you envision the agents being orchestrated?

Marc: A use case would be: this exercise client gets this agent (from the website). She starts to use it; she loves it. I’ll explain an example of what this app could do, because it will trigger contextual stuff. The author has created custom audio for her, so digital media is part of this. When they’re together, the author is talking to her. He’s a trainer he’s telling her, “Come on. Keep going. Keep the spirit going,” and he records that audio. Then, we’re monitoring her performance from the Fitbit, and I can say, “If you’re in a plateau, start Kraftwerk. Or, start some techno. If you stop, trigger the audio, ‘Come on. Don’t stop. Keep going.’” So, that’s contextually triggered.

This woman, who’s now losing her weight, she comes to our site and she sees all these other agents. Some of them are public, like a tour guide or a game master, but some of them are called “life assistant” or “home assistant.” I load the home assistant. Now, I can work with my dropcam and my lights and my smart things and whatever, and I can design something with my house.

It’s going to be really important to understand how to distribute and share that metadata.

Marc: Yeah, but I think: first things first – it’s human-to-human. So, a mom in a house. If I’m managing a house, the first thing I’m thinking about is, “Where is my husband? Is he coming home for dinner? Where are the kids? Have they done their homework? When you all go out, turn off the lights, and when you come back we can turn on the lights.” Again, it’s so simplistic. What can I do that’s fun and exciting with my sprinklers and my candy cabinet?

My favorite lick (‘cause Scoble’s got this dropcam) is: imagine that a home would have a persistent store that’s your list of all the people you interact with – the delivery guys, the nanny, the chauffeur, the gardener. So, Marc Canter comes to the front door: facial recognition – “Oh, it’s Marc Canter” – trigger Bob Marley. If Marc Canter passes in front of the TV set, start Citizen Kane. You can set up these – I call them “bombs,” like when you do a photobomb – you can imagine programming these things just walking around your home and… boom! It triggers something.

This is an example: if I’m a school teacher, I want to create this incredible curriculum. I want to play with these devices and tie them to learning subjects, and actually even be able to hand out a customized curriculum to each of my students.

If I’m an A/V guy and I come to the house, I’m setting up the speakers and I’m getting a hundred bucks to install it. Maybe I’ve gotta little hustle and I come up with a customized app that customizes it for this client, leave it behind – I make 10 bucks a month off of them.

Back when you were starting – multimedia and the evolution of ways to use digital tools, going from analog to digital – that was a renaissance time.

Marc: Totally.

Is this the renaissance time?

Marc: The other renaissance time for expression was blogging. That was some serious shit. With the young people today, being able to interact in Minecraft and do a plugin? Awesome. A mod? That’s awesome. My daughters do that. Being able to do that Snapchat, or that Instagram, and learning about image processing? That’s awesome. Even just making a YouTube video – to have the ability to be a personality and go out there – these are awesome. This is incredible.

What we’re doing is we’re adding the devices. We’re adding proximity; then we can track things indoors. We understand now that even the concept of context comes along. In our tool we have a context map, because the trick here is – remember, I said to get the right authoring paradigm? I use a context map for two purposes. One is to ingest everything. It’s like a fancy setting so I can bring in all these things about the end user or the other people – technology, the real world. I’m bringing all this stuff in – that’s great. But once I do that, I want the context map to tell me when things are changing. And if things are changing, I want you to animate. Because I’m a human, and when things animate, I notice it. “Wow – look: my friend is available for a run.” “Oh shit – it’s going to rain.” “Wow – there’s an event coming up.” All of these things that matter to me as an author – I should be able to paint a picture, choreograph, express my feelings.

You can’t do that if you have to hire seven programmers and spend a million dollars. The Internet of Things really needs authoring tools. In one sense IFTTT is an authoring tool, but what it authors is these simple, little, toy-like connections. I’m hoping that, because they took Andreessen money, they’ll stick in the middleware business. They’ll offer premium service and do analytics, and they’ll take the data and monetize it – let them go do that. I want to recreate what we did with Macromind, 25 years later.

Well, Marc, thank you for taking some time, having some lamb, and we’ll see again soon.

Marc: One more thing I want to say. If you go in and you type in “Marc Canter Theme song,” you’ll come up with an awesome reggae track, okay? So, do that because we love Google so much, and we love YouTube.

We’ll start the show with the audio track.

Marc: Oh, it’s an awesome track. [Singing]


Marc: Thank you, sir.

Thank you.

Feature image via Flickr Creative Commons.

Group Created with Sketch.
TNS owner Insight Partners is an investor in: The New Stack, Island, Persona.
THE NEW STACK UPDATE A newsletter digest of the week’s most important stories & analyses.