06.04.17 Recap

It Could Be 10 Times Cheaper To Take Electric Robo-Taxis Than To Own A Car By 2030

“Is this going to destroy the company? If not, let them test it.”
Love that ethos.  💯   [Also, Reid's @MastersOfScale is pretty great]

Delphi joins self-driving alliance with BMW, Intel

Mossberg: The Disappearing Computer

OpenAI Baselines: DQN

As Profit Dwindles, Ford Is Said to Replace Its C.E.O. with its director of autonomous vehicle research

Self-driving car technology: When will the robots hit the road?

Toyota pushes into blockchain tech to enable the next generation of cars

Zuckerberg Asks Harvard Grads to Fight Isolationism, Nationalism

The Unbundling Of Excel
☝️  I've been hot on this trend for awhile now

This is Jupiter? Giant planet surprises scientists in Juno’s first flyby

China's GSR Nears $1 Billion Deal for Nissan Battery Unit

AlphaGo retires from competitive Go after defeating world number one 3-0

Zillow announces $1M prize for anyone who can improve the algorithm for its Zestimate
Always like to see these open innovation challenges 🙌

When the Patient Is a Gold Mine: The Trouble With Rare-Disease Drugs

Faraday Future Said to Plan $1 Billion Raising as Main Backer Struggles

Facebook to launch ParlAI, a testing ground for AI and bots

Why Google’s CEO Is Excited About Automating Artificial Intelligence

Apple Is Working on a Dedicated Chip to Power AI on Devices

How AI Startups Must Compete with Google

Is China Outsmarting America in A.I.?

I trained a Word2Vec model on FoxNews broadcasts:  This is what it thinks about the world.

 

10.16.16 Recap

Annnd, we're back. Continuing to map various readings, notes, and musings over the weeks.

Differentiable neural computers by Google DeepMind
Blog post on DeepMind's recent work on adding memory to neural networks (see below for paper).

Hybrid computing using a neural network with dynamic external memory by Google DeepMind
The associated paper in Nature from DeepMind mentioned above ☝️.

Graph Powered Machine Learning at Google by Sujith Ravi
Highlights Google's work with graph-based, semi-supervised learning. Pretty awesome work where they are taking multi-graph representations of data and are pairing them with nodes of concepts and edges connected by similarities.

Large Scale Distributed Semi-Supervised Learning Using Streaming Approximation by Sujith Ravi and Qiming Diao
A paper from the Sujith Ravi (author of the Google post above ☝️) that covers some of this work.

The Hive is the New Network by Arjun Sethi
Networks are so 2015, or so they say. Arjun Sethi from Social Capital outlines the notion of a 'hive' and its implications for products and companies.

We don't sell saddles here by Stuart Butterfield
Found this old memo from Stuart Butterfield to the team as they built Slack. Great principles for early stage companies to embrace and hits on core topics from building great product, to selling, to driving change, to execution. TL;DR - Build something people want. Market from both ends. Sell the innovation, not the product (to which he noted, we're not selling "a group chat system...what we're selling is organizational transformation). Who do we want our customers to become? How do we do it? ("We do it really fucking good.") Why? ("Because why the fuck else would you even want to be alive but to do things as well as you can? Now: let’s do this.")...

Read More