In September 2019, I read about New Hand Gesture Technology Could Wave Goodbye To Passwords on Forbes; and in 4 months later in January 2020 Amazon is already working on it, letting people pay with the palm of their hands. Wow you’re fast Amazon!
The power of marketing with an experimental mindset by Think with Google:
- Organization: Centralize versus localize
- Continually testing and learning with new dimensions to find approaches that work — and, equally important, ones that don’t — is crucial to making future campaign decisions based on proven results.
- Prioritization: Maximize impact and scalability
- A good rule of thumb is whether the learnings can be scaled to inform a better strategy for other campaigns without running more individual experiments.
- Testing out different tools, tactics, and strategies is how marketers can uncover opportunities to be more efficient while delivering better, more relevant campaigns.
- Capability-building: Nurturing and sharing learnings. Inspiring an experimental mindset requires nurturing a culture that empowers marketers to test out novel ideas without the fear of failure.
- Test, learn, repeat: Embrace the power of experimentation
The future of customer experience: Personalized, white-glove service for all by McKinsey. This is Service Design 101.
Transforming PMO Into PPO To Become A Product-Driven Enterprise: Interesting articulation by Craig Strong, Principal Advisory at AWS, on moving from Project Management Office (traditional project delivery and output controls, which can constrain emergence and agility if not modernized) to a more business-driven Product and Portfolio Office.
With org design and roles evolving with Digital Business Transformation, the role of Product Owner is fundamentally crucial. Yet, it is often poorly understood in organisations (Business Owner? Delivery Lead? Initiative Lead? Programme or Project Manager? Product Manager or Business Analyst?).
The 4 areas to consider when creating a PPO are music to my ears:
- Establishing Product Teams Who Own Products
- Shifting From Controls To Insights And Triggers
- Technology And Cloud Enablement
- Product Life Cycle And Portfolio Management
GoJek #FirstPrinciples series with Hanna Zia shared Lessons and Practical Advice from Product Management at Google. She worked in Silicon Valley, Switzerland and now in Singapore across hardware and software projects. It was great to see the PM’s responsibilities at Google as you work horizontally without formal authority, the approach to product design, the elements of their PRD (Product Requirements Document) of 10 pages max with Vision, Users, Use cases, Design, Features, Roadmap and Metrics. My favorite part was her illustration of the many hats a Product Manager wears in the dynamic product cycle: mostly Visionary at the definition phase, then becoming a Project and Resource Manager during development (bigger team would have a Scrum master, tech PM or project manager but in smaller teams the Product Manager would be the Scrum master), and moving to mostly a Coordinator and Spokesperson during Launch to Business Analyst during the maintenance. I loved how realistic Hanna was during the presentation, sharing that no one reads PRDs: she advised in known environment to amend an existing PRD, find a peer reviewer to bounce off, talk to engineers & UX lead to ensure they support the PRD and then only plan a cross-functional team with backlog, MVP, roadblocks etc.. before starting work. For a new idea, she recommended creating a 1-pager that shows the main user journey (ie. everything goes well) and what user will do. Lots of 1-pager fail so fail early and often, then iterate and talk to engineers, UX and start the PRD as a Google doc so it can evolve, with visual sketches, mocks or prototypes. You don’t have to think about all the edge cases yet. When you find new information not covered in your PRD, write up your reasoning on a 1-pager and update the PRD after alignment with leadership and over communicate with your cross-funtional team which includes legal, marketing, BD…
Vision without execution is hallucination.
I loved Hanna’s perspective on shipping vs landing: launching is not enough, you have to land it! Dog fooding means that we need to eat our own dog food, we use & monitor our product and think of improvements or fix bugs. Her philosophy on executing is spot on: Scrum deals with a lot of unknowns, PMs do their work at the same time as engineers and UX and the PM role is to provide clarity. As a PM, you have to be a jack of all trade eg. learn Sketch, Figma even if you have UX designers as they are busy on the next launch. Your goal is to unblock and be fast so you often help with wireframes, quick decisions. You do whatever it takes, have the least ego. You’re the biggest advocate for your product to get more resources and remain humble. You’re a service provider to ensure everything gets out of the door, but you’re in the background, you give lots of credit to others. You take responsibility to fix features that are not working, you sweat the details and remain optimistic.
This is so exciting (my review of GPT-2 last year is here). A16Z podacast’s 16 Minutes on the News #37: GPT-3, Beyond the Hype has it all. Of course GPT-3 was trained off of 175 billion parameters from across the internet (including Google Books, Wikipedia, and coding tutorials); its code contains bias. See my view on technology and our human bias.
Yet, it’s good to know in Wired and 3 examples that less than a day after Facebook’s head of AI Jerome Pesenti called out bias coming out of a program created with GPT-3, OpenAI launched a toxicity content filter API, which rates all content created by GPT-3 on a toxicity scale from one to five, and anything above a two is flagged for moderation.
“There’s doubtless a lot of biases we haven’t even noticed yet,” says Anders Sandberg, a senior researcher at Oxford University’s Future of Humanity Institute. “It wouldn’t surprise me if we started to use systems like this as tools to detect the weird biases we have.”
“The paradoxical thing is that these text systems actually are pretty good at calculating the probability that something was written by them,” says Sandberg. So rather than helping to stimulate troll factories, GPT-3 could keep its own fake news in check.
Another example on Twitter: =GPT3()… the spreadsheet function to rule them all. Impressed with how well it pattern matches from a few examples. The same function looked up state populations, peoples’ twitter usernames and employers, and did some math.
The MIT technology review reported in July that iGPT, based on GPT-2 was swapping words for pixels and train the same algorithm on images in ImageNet, the most popular image bank for deep learning. Because the algorithm was designed to work with one-dimensional data (i.e., strings of text), they unfurled the images into a single sequence of pixels. They found that the new model, named iGPT, was still able to grasp the two-dimensional structures of the visual world. Given the sequence of pixels for the first half of an image, it could predict the second half in ways that a human would deem sensible. I wonder what happened to the butterfly in the elephant ears 🦋 🐘 👂 🤔
I love revisiting predictions throughout the year: 2nd Jan 2020, these 10 predictions for the 2020s missed quantum. But these 8 Predictions for Quantum Computing in 2020 listed as #3: Quantum Innovation Will Blossom in 2020.
“Amazon Braket enables FCAT to develop hardware-agnostic software so we can easily switch to new quantum systems as they become available. We’re able to research the strengths of different quantum backends and build hybrid classical-to-quantum and quantum-to-quantum workflows.”
Microsoft announced in May 2019 recent partnerships as an update on its quantum computing services division Azure Quantum. Not a quantum computer yet but paving the way for the next revolution. Toyota Tsusho is using Azure Quantum for traffic optimization and other mobility-service experiments.
If that doesn’t scream innovation, I don’t know what does.
To recap on Quantum, Quantum Computing Expert Explains One Concept in 5 Levels of Difficulty
WIRED has challenged IBM’s Dr. Talia Gershon (Senior Manager, Quantum Research) to explain quantum computing to 5 different people: a child, teen, a college student, a grad student and a professional.
And it’s amazing!