Suggestions
Paul Viola
Distinguished Engineer at Microsoft
Paul Viola is a Distinguished Engineer at Microsoft. With a Ph.D. from MIT, he holds 35 patents in advanced machine learning, web search, data mining, and image processing, and has authored over 50 academic research papers.
Paul's career spans various roles from a Professor of Computer Science at MIT to being a Computational Scientist and Hacker at Self, VP Science at Prime Air, and Chief Scientist at Highspot, Inc. His expertise includes developing highly engineered algorithms used in digital cameras, surveillance, medical imaging, and surgical planning.
At Microsoft, Paul initially worked as a researcher applying machine learning to document analysis and later founded Bing's research efforts in machine learning and text understanding. His leadership led Bing to achieve search relevance comparable to Google and significantly increased Bing Ads revenue.
Paul Viola's educational background includes a Ph.D. and S.B. in Computer Science from MIT. His organizational affiliations include positions at Amazon, Highspot, Inc., and various roles at Microsoft.
Paul is recognized for his contributions in machine learning, search algorithms, and image processing, with a career that exemplifies excellence in research, engineering, and leadership.
Highlights
Hey LinkedIn braintrust. There is so much noise flying around the recent advances in LLMs and if they are AGI (or on the road to AGI, or a dead end).
Anyone know of a nice reference to the more practical issues? Sort of at the Scientific American level? Perhaps a list of 5-10 articles would do the trick. Or a very recent book.
I have had a number of friends ask. The stuff I read is not what they would find helpful.
My previous post was about how #spacex and #tesla seem to have achieved so much, so rapidly, with few built in advantages. I know nothing about how they run internally. Anyone have some wisdom to share??
There is accepted wisdom about what makes a productive workplace. Collaborative teams. Supportive and safe environments. Inclusion.
I think that if you are trying to get the most out of your existing workforce this can be a big deal. You are essentially trying to substitute loyalty to the team (which is natural for humans) for loyalty to the vision or the product. Most products don't exactly attract a lot of loyalty... "we are going to create the best darn laundry soap ever!". But many people can love their team.
Being supportive, collaborative, and inclusive is important for kids sports teams. There is no way to keep them engaged if you don't give them all positive feedback, or if some kids never get to play. But there is a huge transition as teams get more competitive (say high school teams that want to win the state championship). In that environment folks are given roles based on their ability to contribute (though it is often overlooked that a good player makes the players around them better). Its a tough transition for many kids, and many don't make the cut (like me!). But if you can't score or defend then you may not stay on the team. Of course there are "no cut" teams and recreational teams. Good fun, but no championships.
I suppose you could ask if your employees are more like the young kids or the high school champs.
I have only been in a few scrums (and of course there are many ways to run a scrum). But I feel the biggest value is that the team stands up and looks at each other every morning. Everyone presents progress. And it is hard to come in for 3 days straight and not have made progress. A good scrum master will quickly ask: "do you need help?", "should we transfer the workitem to someone else?". Scrums are not confrontational (if done well). But letting down the team, publicly, in the scrum is a powerful motivator.
Friendly competition helps as well. I remember working with collaborators that could get a ton of work done in 24 hours. I quickly realized that I wanted to get just as much done. And that gave me additional motivation.
What is the most productive environment folks have worked in? Not the one that was most fun/easy/relaxing (though it could have been).
Why was it productive?
How can we setup processes and mechanisms that create these environments?
