Data & Tech Stuff, Vol. 2

Phew -- what a month!

I've only made one single 2020/Hindsight joke this year so far, which, in hindsight...

OK let's get to it!

As always, email me any thoughts you have or links you want thrown in here. And please feel free to forward this newsletter to anyone you think might be interested.

I, for one, welcome the AI Winter...

There have been more recent reports (BBC, Forbes, HackerNews) that we may be on the cusp of an “AI Winter,” where the pace of innovation, advancement, and implementation of artificial intelligence (AI) slow down compared with the “Renaissance” we've seen in recent years.

I'm a little of the mindset that all of this talk is relative -- because what are we even talking about when talking about AI, anyway? It's not like we're up to the level of artificial general intelligence quite yet -- so what is slowing down? Essentially, there's a recognition by folks far and wide (investment to government to public interest) that the enthusiasm and supposed benefits of an AI world aren't all they're cracked up to be. The algorithms are biased, not transparent, and not ready for adoption in prime time. This isn't just an academic point--three cities have already banned facial recognition and surveillance implementations, a technology that relies heavily on AI.

I think one of the challenges is that the term “AI” has become a placeholder for technological magic – in much the same way that big data or business intelligence or mobile once were. We should update Arthur C. Clarke's Third Law to say, "Any sufficiently advanced technology is indistinguishable from AI."

I'm not an AI hater, it should be noted. I should probably disclose here that I'm someone who spends a fair amount of time in the workplace trying to figure out when/where to apply new and emerging technologies for use in projects — solving tough challenges with the right mix of people, information, and technology.

This is all to say that not only am I not worried about an AI winter, but I actively welcome it. Let me explain why.

I keep coming back to that now-daily mantra of "AI in the pitch deck, machine learning (ML) in the project plan, and linear regression in the execution." (if Python is more your speed) I'm not saying that everything we see around AI is lipstick on a pig -- just that there is a LOT of hype out there, and for every one legitimate innovation with AI/ML, there are about 100 of "probably just analytics on a spreadsheet" under the hood. Conservatively.

Too much time seems to be spent on knocking down false claims about what's possible with current constraints in ML, where it would be best to apply particular frameworks and novel technological solutions over traditional methods, and where we should be prioritizing our limited resources as organizations working in the research-meets-social-impact-meets-technical-assistance-meets-sustainable-development arena.

This isn't to say we are slowing down our interest in what's on the horizon, by any means. My team is particularly excited about a current project focused on crafting a real-world standards checklist for ethical AI in our programs, for example (and building off what we're seeing from the White House on AI), and taking some of our recent lessons and packaging them into a more universal and accessible how-to guide for AI projects in social impact settings more generally.

We don't always need an AI Renaissance to be driving full-tilt — sometimes it's nice to have an AI Winter while we have a Sustainability Spring or an Implementation Summer (™, obviously).

A “Technology for Good” Roundup

• Development Informatics - Global Development Institute - The University of Manchester — I hadn't seen this list before -- absolutely monumental list of working papers with ICT4D (Information and Communication Technologies for Development) focus. Something for everyone in here -- digital folks, health folks, ag folks, development folks, you’ve got it.
• Do no digital harm - Tech’s Good — The Digital Principles of Development are a good start, but they aren't perfect. The global tech space has been missing something akin to medicine's “do no harm” Hippocratic oath — is this it?
• USAID Digital Investment Tool October 2019 — This is a *really* solid framework for approaching digital within projects (particularly when you're starting the conversation with stakeholders).
• Technical Literacy learning tracks to boost nonprofit digital skills — The TechSoup Learning Track is interesting enough in its own right.
• Responsible Data for Children (RD4C) — A joint endeavor between UNICEF and The GovLab at New York University seeks to highlight and support best practices in data responsibility. Great guide regardless if you're working at the intersection of data responsibility and social good.
• 2019 Global NGO Technology Report — My takeaways (tl;dr infographics) -- everyone has the same problems you do. Reassuring/terrifying.
• Building digital medical records for floating villages in Cambodia — Also, the idea of a "local tech whisperer" is **genius.**

Tidbits

• 40-year-old tractors are now a hot commodity - Marginal REVOLUTION — Whoa. DRM limitations and tinker-ability. This development has it all.
• 23andMe Licenses Drug Compound to Spanish Drugmaker Almirall - Bloomberg — Didn't see that one coming. 🙄
• Africa is the fastest growing continent for developers globally — Quartz Africa — Anecdotal evidence backs this up.
• A Drone's-Eye View of the Quick and the Dead  - Scientific American — Machine learning in disaster response.
• Ocean Cleaning Device Succeeds in Removing Plastic for the First Time - EcoWatch — Let's not bank on it…but still…cool.
• This AI kitty litter box will analyze your cat’s poops for some reason - The Verge — Nevermind -- there 👏 is 👏 no 👏 AI 👏 Winter 👏.

Thirsty for more? Check out what else we are doing and talking about when it comes to Digital Transformations.

Cheers,
Gabriel