Notes from 17–24/8

COVID-19

Automated and partly automated contact tracing: a systematic review to inform the control of COVID-19

Taken together, the modelling studies that we identified showed that the effectiveness of automated contact tracing in reducing disease transmission depends on both population uptake (eg, of contact-tracing apps) and timeliness of intervention (eg, quarantining contacts).

As with manual contact tracing, automated contact tracing also relies on accurate and reliable identification of encounters during which transmission occurs.

Academics have warned of the risks that automated contact tracing could pose, including a so-called mission creep towards unprecedented surveillance and eroded public trust, should data be misused or hacked.

Privacy and information governance are also highly important within manual contact-tracing systems, but given the substantially larger amounts of personal data (including colocation or location data, or both) that could be collected and processed in automated systems, they are particularly important considerations in this context.

Decentralised automated contact-tracing systems benefit from Apple and Google’s support, meaning that interoperability between countries with such apps is likely to be more straightforward than between countries that use centralised systems.

However, a study reported that centralised systems assess transmission risk more accurately (reducing the number of people quarantined), enable better optimisation, are less susceptible to false reports, and are more readily evaluated.

Blockchain

What Is Blockchain Technology?

An analyst’s perspective

Career

Two Books That Turned a 26-Year-Old Programmer Into a Billionaire

“Influence” by Robert Cialdini

1. Reciprocation refers to the human need to return the favor. For example, Hare Krishna members always give passers-by a flower before asking for donations — this makes them more likely to donate.
2. Consistency refers to the human need to “save face” in front of others. In human societies, we respect people who do not change their stance. So, for example, if you build personal rapport with a prospect by genuinely asking how they’re doing, they’re going to feel motivated to consider your offer later — simply because they’ve already been nice to you before.
3. Social proof is the bread and butter for entrepreneurs. All animal species have largely evolved via imitation — long before articulate human thoughts were born. That’s why TV shows use canned laughter, and why bartenders put a few dollars into the tip jar before the customers walk in.
4. Liking refers to an old truism that “we do business with people we like.” Here, the author refers to less-revolutionary truths that we often make business decisions based on someone’s looks and familiarity with us.
5. Lastly, the author reminds us that authority plays a major role in business. Highlighting things like titles and money may seem wrong and inaccurate to a logical person, but they do work on a subconscious level. For example, someone introduced as a professor seems taller than someone introduced as a graduate student. Also, car drivers wait longer before honking on a luxury car than an old car.

“High Output Management” by Andrew Grove

1. Sales forecasts answer the question: how many customers are you expecting to serve today? This has direct implications on how many staff and inventory do you need to keep the customers satisfied.
2. Inventory levels seem like an obvious parameter to watch but is so often forgotten by inexperienced managers. CEOs who fail to evaluate their inventory levels fail to fulfil orders, and that is about the worst customer experience there is.
3. Condition of the equipment refers to your production instruments. As a waiter, have you checked whether the toaster is working today? The customers won’t care if “it’s not your fault.” They’ll simply choose another place next time.
4. Workforce indicators are needed to keep track of your staff. Just like it’s your responsibility as a manager to make sure the toaster is working, you need to check if any of your kitchen staff called in sick today. In that case, you’ll need an urgent replacement.
5. Quality indicators are often forgotten by managers. Did the customers like the breakfast? Will they come back again? Companies that fail to learn go bankrupt, and Tobias learned this lesson without having to fail himself.

Data

How I’d start learning machine learning again (3-years in)

The advanced path (6–12+ months/ongoing)

Once I’d gotten some foundational machine learning skills, I’d build upon them with the following.

• All of fast.ai’s curriculum(s) — practical use cases of many deep learning and machine learning techniques. Watching one fast.ai lecture turned into a solution we built for a client.
• Any of deeplearning.ai’s curriculum(s) — choose the one which sparks your interest the most. Compliments fast.ai’s practical approach with theory.
• Full-stack deep learning curriculum — this is where you’re going to tie together the machine learning knowledge you’ve got with the web development knowledge you’ve been learning.
• Replicate a research paper (or multiple).
• Hands-on Machine Learning Book with Scikit-Learn and TensorFlow Part 2 — TensorFlow focused but the concepts bridge to many different applications.

Development

How To Fetch Data From Google Sheets with React and Tabletop.js

As developers, we’re constantly searching for ways to improve our everyday workflow. Here’s a way to save time by fetching data straight from Google sheets with React and tabletop.js.

Finance

Venture Capital’s Role in Financing Innovation: What We Know and How Much We Still Need to Learn

One promising response to these challenges is to rethink the organizational model for incubating and financing “tough tech” ventures. The venture approach entails entrepreneurs coming to venture capital firms to pitch them new ideas and the firms deciding whether to fund them. This approach has the benefit of enabling the investors to maintain an arm’s-length relationship from the entrepreneurial team, reducing the entrenchment that is sometimes associated with corporate research and development and internal capital markets.

An alternative approach that has begun to be used by some venture capital investors specializing in biopharmaceuticals (such as Third Rock Ventures and Flagship Pioneering): to incubate and finance ideas in-house. This process has the benefit of reducing asymmetric information because much of the staff for the team of entrepreneurs comes from within the fund. It also enables the venture capital firm to fund what it might believe is the most promising idea or approach, as opposed to selecting among the ideas that walked in the door.

It is also natural to wonder whether collaboration with other parties — governments, non-profits, and corporations — might alleviate some of the barriers to financing new ventures in more difficult technologies. But the track record of these collaborative efforts has been quite mixed. There have been successes, such as the Israeli government’s jump-starting of its venture industry through the Yozma program that leveraged public money to attract private investment, or the success of many pharmaceutical firms in responding to the biotechnology revolution through their venturing initiatives. At the same time, anecdotes abound of naïve officials making poor decisions. For instance, the leadership of Boston University of put one-third of the university endowment into a single faculty-founded biotechnology company, Seragen, an investment that was ultimately sold for pennies on the dollar

Product

The Zendesk Triple Diamond

Use Case Storyboarding

Concepts — Examples looked like:

• Ease of finding music that matches user’s style.
• Communications are context-aware.
• Traveler is able to find destination with more confidence.

Activities — The general form is <role> wants/needs to <activity> which might turn into:

• Mechanic wants to understand <specific/general> problem.
• Designer needs to find a <style/mood/font/…>.
• Data Scientist wants to <ETL/visualize/process/upload/…> data.

More complex examples might be formed like:

• Conversation started by business owner asking <something relevant> about a <object> via <medium>.
• Trip started by passenger asking <something relevant> about a <destination> via <medium>.
• User looking for <style> music by <feature>.

For an activity, what is the beginning…

How does the user enter the situation? What does the user have on hand, physically, mentally, or on-screen?

1. User is at their computer and receives a message where they don’t know/understand…
2. User is looking at a video and…
3. User is out and about and manager calls them…
4. User checks their chat to see if anything important has come up on order <XYZ>…

What about the activity do you want to highlight…

• User goes to <what> and starts chat and…
• Instead of needing to <X>, user initiates chat within the <context>…
• User sees service technician is not present for chat, but can ….

Chain together as many activities as seems appropriate and realistic.

The Hierarchy of User Friction

GIves some salient examples of how they were solved

Product Vision FAQ

The product vision is the explicit responsibility of the head of product, which in startups and growth-stage companies, is often one of the co-founders. This is not something that is delegated to individual contributor product managers or product teams, although they can of course contribute and review.

These are the factors that I consider essential as we craft the product vision:

• Business Objectives and Constraints
• Customer Problems
• Critical Insights
• Enabling Technologies
• Industry Trends
• Competitive Landscape
• Go To Market Considerations
• Our Own Capabilities
• Organizational Impacts

Open Source

Producing Open Source Software

Book + talks, perfect

Tools

Dropbase Centralize offline data. Import files. Process and clean up data. Export to a live database with 1 click.

Might be useful someday