Cognitive OpenTech

Whether we know it or not, and most of us don’t, we will all soon be building our own personal cognitive systems.   Nearly everyone is on a journey as a developer/maker/citizen scientist/data wrangler/lifelong learner/T-shaped service scientist…  people with depth for problem solving and breadth for empathy and communication skills, evolving to work better on multidisciplinary, multisystem, multiculture teams to take on larger and larger challenges.

We will build them, and they will know us well, and keep our data private.  We will be able to build them, because the building blocks are getting better fast.   For example, just sixty years ago it required the combined resources of a nation to put a satellite into orbit. And now it is rapidly becoming something that a single teenager can aspire to do – from DIY satellite to DIY rocketry.  Making the world safe and secure, and escaping fear of harm from others is one of the biggest challenges we face.  Making tools to solve our most pressing problems is getting easier and easier as the building blocks get better – rebuilding society from scratch is becoming easier, not harder, to imagine.

Cognitive Open Technology building blocks (Cognitive OpenTech) are getting better, and they include a wide range of capabilities as a service: Augmented Intelligence, Artificial Intelligence, Machine Learning, Deep Learning, Exoskeleton Clothing, Maker/Rapid Rebuild Tools, etc.

However, there are still many unsolved grand challenges remaining: http://service-science.info/archives/4447

These challenges are best approached as an open ecosystem of collaborators, working on technology as a service – see https://www.partnershiponai.org/

Regarding Open Technology – if you have 90 minutes, you really should watch this:
https://www.youtube.com/watch?v=h9FSqVbdHis

Agenda

0:00-2:00
Intro animation

2:00-3:00
Sam Ramji (Google, VP Product Development – Developer and Compute Platform)

3:00-7:00
Sarah Novotny (Google, Program Manager – Kubernetes Community Development, Google Cloud)

8:00-15:00
Vint Cerf (Google, VP and Chief Internet Evangelist)

15:00-22:00
Jim Zemlin (The Linux Foundation, Executive Director)

22:00-31:00
Eric Brewer (Google, VP Infrastructure)
(26:00 – Thousands of Googlers contributed to 1000’s of open source projects in 2016)

31:00-40:00
Chris Wright (RedHat, VP and Chief Technologist)

40:00-47:00
Jeff Dean (Google, Senior Fellow)
(4000 projects use TensorFlow insider Google today)
(I think this means: Google promotes “open” as key to employees, customers, developers, researchers, universities, and citizen science. The distinction between employees and customers may decrease over time as everyone becomes a developer/lifelong learner)

47:00-55:00
Rajat Monga  (Google, Principle Engineer)
Examples including skin cancer detection and eye disease:
http://www.nature.com/nature/journal/v542/n7639/full/nature21056.html
http://jamanetwork.com/journals/jama/article-abstract/2588763

51:00-1:04:00
Sam Ramji (Google, VP Product Development – Developer and Compute Platform)
“Best time in history to be a developer – craft is respected”
“Age of open source – fast innovation.”
“Days of committing to a closed ecosystem are over”

1:04:00-1:05:00
Alison Wagonfeld (Google, VP Marketing – Google Cloud)

1:05:00-1:14:00
Will Marshall (Planet, CEO)
Satellites to map the surface of Earth.

1:14:00-1:19:30
Willem Sundblad (Oden Technologies, CEO)
Empowering manufacturers with data

1:19:30-1:26:00
Herman Narula (Improbable, CEO)
SpatialOS – Simulation of cities for policy making, game development platform

1:26:00-1:32:00
Alison Wagonfeld (Google, VP Marketing – Google Cloud)
Anthony Goldbloom (Kaggle, Co-Founder and CEO)
Scaling up to deep learning from simpler machine learning

1:32:00-1:30:00
Alison Wagonfeld (Google, VP Marketing – Google Cloud)
Santi Subotovsky (Emergence Capital, Managing Partner)
Matt Ocko (Data Collective, Managing Partner)
Last ten minutes is VC funding for microservices on cognitive open tech…

Fei-Fei Li (former Stanford, Chief Scientist, Google Cloud for AI)
https://www.youtube.com/watch?v=Rgqgdddl018

11:00 “…democratizing compute, algorithms, data, models, talent and expertise.”
12:00 “…model for AI… best papers for AI competition – quickly turned into products and services.”
13:00 “vision API and meta-data from Google search to recognize entities in Google Knowledge Graph”
14:43 “video-intelligence API – Sara Robinson, scene changes in video”
Sheep dogs and mops – https://www.youtube.com/watch?v=w1xNTLH1zlA

These are also important:

OpenWhisk for serveless microservices

BonnyCI for rapid rebuilding

Jupyter Notebook and used by DSx (Data Science Experience)

Sonnet is open now too.

CIO’s are using AI and Open Technologies to continuously improve their organizations:

CIO: https://dupress.deloitte.com/dup-us-en/focus/tech-trends/2017/moving-beyond-artificial-intelligence.html

Go Nat Friedman – $5K grants for Open AI projects: https://nat.org/aigrant-get-5-000-for-your-open-source-ai-project-1118dd7db083

Open source AI/ML/DL for smarter/wiser service systems is a passion, including circular economy and rapidly rebuilding from scratch for sustainability and resilience of systems – helping all nations become energy independent is a key first step, and geothermal needs to be 100x cheaper than today, but that is a good challenge to work on….