When you start a company you wear all of the “hats.” As the entrepreneur, you are the human resources department, the chief of finance, the lead developer and the janitor. Your goal as an entrepreneur is to find people for those jobs that are better at them than you are. When you find a stellar developer, you hire them and give them the developer hat. Same for HR and for finance. If you do your job right, at the end of the process the only hat you’re wearing is the one which fits you best.
Today I’m happy to announce that I’m taking off the CEO hat and passing it to Michael Lewis, who is significantly more qualified to wear it.
Michael Lewis is the founder of Stellar Semiconductor which was sold to Broadcom in 2000 for $161M, then went on to found Cryptic Studios which was sold to Atari in 2008 for more than $26M. For the past decade Michael has been deeply involved in charitable efforts through the Anspar Foundation (formerly The Lewis Charitable Foundation).
Open-source software is a fundamental interest of the Anspar Foundation. In keeping with the Foundation’s focus, it invested in Mycroft in October of last year and is on track to become the largest investor in Mycroft AI Inc.
Michael is joining Mycroft as our new CEO, which will allow me time to work on the technology for the first time since I hacked together v0.0.1 in March of 2015. But I will continue to be involved at the executive level and work closely with Michael to maintain our ethos. My new title, inspired by Jonathan Frakes of Star Trek fame, is “First Officer.” It may be a bit NAVY, but it’s also accurate. Co-pilot seemed a little odd.
I’m really excited to have someone as experienced as Michael join the team. I’m already learning a lot about focus from working with him and am looking forward to collaborating going forward.
So without further adieu: Mr. Michael Lewis.
It’s a pleasure to say hello and have the opportunity to tell you a little about my story.
When I left Cryptic Studios I created a private foundation to pursue some ideas I had about transparency and openness in business that I had formulated as a result of some of the less satisfactory aspects of my success. Through this foundation, I’ve funded research and even embarked on some research of my own (opengamelabs.org). I’ve backed a lot of open hardware and software projects. I’ve followed the privacy issues of the major platforms with interest and dismay and even took a foray into the concept of internet identity and reputation networks. But when I looked at Mycroft and discussed it with my colleagues a lot of pieces clicked into place. The work Mycroft is doing touches on all of these issues and more.
Conceptually, “voice technology” is no longer novel. But we’re still at the beginning of this story: The potential for voice technology is far larger than any of my previous projects. The roadmap ends with “ubiquitous voice control everywhere”… and that has a lot of implications. Whether it is achieved by giving away your life experiences to one of a handful of monopolies or in a way that leaves you in control depends on whether Mycroft and companies like it succeed in inspiring enough people to believe that a privacy-respecting, user-first system is possible.
I think the world needs Mycroft. I have no doubt that we can make a commercially successful product that nets us a worthwhile financial return. What is exciting for me is the prospect of doing something more. So I’m coming out of my cushy quasi-retirement to add my experience to this team, to continue to learn and gain new experiences in an environment steeped in the ideals I’ve been engaged with for the past decade. And, maybe, just maybe, together we can change the world for the better. At least a little bit.
While it is true that the Mark I hardware has seen continuous updates over the years, the support and QA for that effort has been shrouded in lore, requiring a dedicated team of specialists with a “need to know” level of unintentional secrecy. I advised the team in December that they should focus on process discipline and QA: establishing regression tests, ensuring repeatable results, documenting how things work and rigorously tracking bugs and tasks. The results are already noticeable and the groundwork is laid for the next step, which is a well-defined, readily supportable, minimum viable product. Our first new product target is an end-user installable software package that doesn’t require any technical expertise. In addition to providing customers (from enterprise to home-user) with a better experience, this in turn becomes the framework in which we can make meaningful progress on improving the machine-learning algorithms and other features. It is also a prerequisite to developing a system with well-characterized real-time requirements for which we can properly specify cost-effective hardware. To be clear, this is all incremental work on top of the many person-years of development work that has been done. The current team of just four developers has made great strides on this path, but we will need to increase the staff along the way to shipping commercial products.
I’ve enjoyed working with the Mycroft team in an advisory capacity over the past few months. Every one of them is making exceptional contributions to the business and software, and it is a great core team on which to build. It is too early for me to make specific promises involving timelines, but we’ll keep you updated as we refine the roadmap I’ve outlined above and we will continue to keep the communications with the community flowing freely.
I look forward to the journey ahead of us and welcome your questions and feedback.
Michael C. Lewis