All posts filed under “Machine Intelligence

comment 0

@techvitamin 2.5: Dave Cotter, CEO of Reply Yes

Dave Cotter is one of those people who is deceptively sunny. Sure he's engaging and funny, but you figure out pretty quickly that he's juggling something complex, really ambitious, and that he has a deep pool of that essential founder's gift: faith. He's no babe in the startup woods, and is an Amazon and RealNetworks vet. Reply Yes, Dave's current venture, is a mix of retail savvy and messaging and AI, and is at the center of what is being called "conversational commerce". They've launched two messaging centered "stores" -- The Edit, for vinyl records, and Origin Bound, for graphic novels -- where the simplicity of the offering belies a tremendous amount of tech and logistics and painstaking attention to the customer.

Dave Cotter, CEO of Reply Yes

Sure Dave Cotter’s engaging and funny. But you figure out pretty quickly that in Reply Yes he’s juggling something complex, really ambitious, and that he has a deep pool of that essential founder’s gift: faith. It’s not denial, just bedrock confidence. No babe in the startup woods (he was a co-founder of SquareSpace), he’s also done the larger company thing at Amazon, Zulily and RealNetworks.

Dave’s current venture is a mix of retail savvy and messaging and AI, and is at the center of what is being called “conversational commerce”. Inspired by the sheer simplicity of text, and to some extent by what’s been going on in China with WeChat’s platform, Reply Yes — and a host of other startups (Magic, x.ai, Peachd.com, etc.) — has been running hard at this problem for the past few years.

Dave’s team has launched two messaging centered “stores” — The Edit, for vinyl records, and Origin Bound, for graphic novels — where the simplicity of the offering belies a tremendous amount of tech and logistics and painstaking attention to the customer. The company is a product of Madrona Venture Group’s labs, and in December raised $6.5M in a Series A — bringing their total to $9M.

In this episode we talk about what he’s done to get Reply Yes going, how they’ve managed to focus, and navigate the crazy world of music brands, while fundamentally innovating at the edge of natural human interfaces and offer personalization. How does he as CEO make tradeoffs between going very deep in vertical specialization, yet keeping an eye on the big platform play?

While Reply Yes came out of the gate with a text message centered product, since that time Facebook, Microsoft, Apple and others have made messaging — and in particular transacting over messaging — a much greater focus. We talk about how they are navigating this…dancing with goliaths, but taking advantage of the tech they are providing at the same time. A classic entrepreneur’s dilemma.

Have a listen.

Play
comment 0

@techvitamin 2.4: Soma Somasegar, Venture Partner

If one heard that someone had spent 27 years at Microsoft and then left to spend time investing in startups, you could be forgiven for thinking that it was really a form of retirement, and an opportunity to dabble. But Soma Somasegar -- who's last position at Microsoft was the Corporate Vice President of the Developer Tools division -- doesn't come across as content, or playing, or, well, done.

Soma Somasegar (Photo courtesy of Geekwire)

If one heard that someone had spent 27 years at Microsoft and then left to spend time investing in startups, you could be forgiven for thinking that it was really a form of retirement, and an opportunity to dabble.  But Soma Somasegar (@SSomasegar) — whose last position at Microsoft was the Corporate Vice President of the Developer Tools division — doesn’t come across as content, or playing, or, well, done.

In this episode we talk about his big career switch, the white hot battle in cloud computing between Google, Microsoft and Amazon Web Services, and how entrepreneurs should think about their tech stack choices.  We also talk about how Microsoft will stay relevant for another generation of developers, including embracing Linux, Python and even putting Visual Studio on MacOS.

Not surprisingly, he’s very bullish on AI, and has some interesting thoughts about how it will manifest, how humans will stay relevant, and how the different players will play to their strengths. He also talks about Madrona’s investment framework on AI and Machine Learning, and some of their experiences with Spare5, Dato/Turi (acquired by Apple), and Kitt.ai.

Have a listen.

 

Play
comment 0

@techvitamin 1.7: Ed Lazowska, Bill and Melinda Gates Chair in Computer Science

Are our Tesla's going to band together in the Costco lot and attack us? Find out. In this episode Ed Lazowska, the eminent and long-time member of UW's Computer Science faculty joins Michael Cohen and I to discuss everything from big data, deep learning to how Universities are responding to the massive demand for computer savvy graduates. Ed's brilliant yes, but also a very animated and entertaining story teller.

Ed Lazowska, Bill and Melinda Gates Chair in Computer Science & Engineering

Are our Tesla’s going to band together in the Costco lot and attack us? Find out.

In this episode Ed Lazowska (@lazowska), the eminent and long-time member of UW’s Computer Science faculty joins Facebook’s Michael Cohen and I to discuss everything from big data, deep learning to how Universities are responding to the massive demand for computer savvy graduates. Ed’s brilliant yes, but also a very animated and entertaining story teller.

If you have children, you’ll want to listen because UW (and NYU and Berkeley and others) are doing cool things to bring computational/data fluency to programs far beyond STEM.

Other topics: the difference between knowledge and understanding in AI, and whether programmers even understand the decisions their creations are making. Crazy stuff to contemplate, and Michael and Ed are in the center of it.

A bit shorter because Comcast melted down about 2/3rds of the way in, and we lost a bit of good audio…

Play
Blaise leads a team at Google focusing on Machine Intelligence for mobile devices—including both basic research and new products. His group works extensively with deep neural nets for machine perception, distributed learning, and agents, as well as collaborating with academic institutions on connectomics research. Until 2014 he was a Distinguished Engineer at Microsoft
comment 0

@techvitamin 1.1: Google’s Blaise Aguera y Arcas on Machine Intelligence

Blaise Aguera y Arcas talks Machine Intelligence and inventor of Photosynth

Blaise Aguera y Arcas (@blaiseaguera)

In this first episode of @techvitamin, Blaise Aguera y Arcas (Google) and Michael Cohen (Facebook) join us to discuss Machine Intelligence (MI) across a broad range of subjects, including it’s impact on art (and the impact of technology in general on art), how MI research should be funded, the collaboration between Academia, Business and Government, and much more.

Can machines now create art, independently of humans? Blaise and Michael talk a bit about DeepDream, and the resulting images’ similarity to those conjured by the human brain (perhaps just a tad under the influence). Here’s one:

Animated DeepDream image courtesy of @samim and Github, from Blaise Aguera y Arcas talks machine intelligence

Courtesy GitHub/@samim

We touch on the recent defeat of a Go world champion by Google artificial intelligence: “…the last nail in the coffin of games being an indicator of human intelligence.”

Blaise leads a team at Google focusing on MI for mobile devices—including both basic research and new products. His group works extensively with deep neural nets for machine perception, distributed learning, and agents. Blaise is a well-known speaker on subjects ranging from digital photography to mapping. He’s given three TED talks: on Seadragon (a company he sold to Microsoft in 2006); on Photosynth, which he invented at Microsoft, and Bing Maps.  He’s gave a talk at WIRED2014 entitled “The next big frontier is the mind and the brain.”

Before Google, Blaise was a Distinguished Engineer at Microsoft, where he worked with Michael, who is one of our group of rotating @techvitamin co-hosts. Before his current gig as image and video guru at Facebook, Michael spent 21 years in Microsoft Research, and is one of the world’s foremost thinkers on computational photography (aka, getting photons turned into bits). Both Michael and Blaise were at Princeton (as faculty and grad student, respectively).

We’re always looking for community feedback, so feel free to comment below (or on Twitter, FB, or your preferred vehicle). And please, LIke/Share/(re)Tweet to your heart’s content. We’ll steadily improve the audio (all participants are remote from each other, so we’re a little dependent on mic quality, Skype clarity, etc.). We have a good outtake of Blaise scrambling to find another room at Google HQ after getting kicked out of the one he was in. And, we’ll tweak the show — to a depth and length that makes sense. For instance in this episode, we probably could have talked for much longer, but chose to cut things off to keep it under an hour. Might make sense to keep it going. We’ll get better.

1Blaise mentions a study about government research dollars seeding much of the technology in the iPhone. He was thinking of work by Mariana Mazzucato.

Play