This is the monthly update from Memia Labs; for more information about any of this, please contact them at labs@memia.com
First And Foremost, AI
Lots going on as usual in the AI field this month.
Google announced their new PAIR – People + AI initiative to advance research and design of people-centric AI systems.
Deepmind taught its AI agent to do Parkour using reinforcement learning
The European Parliament debated Do Robots Have Rights?
Elon Musk reprised his doomsaying about AI Safety at a meeting of US State Governors. (In follow up I got interviewed on NZ’s TV3 news in my role at the AI Forum NZ).
While Musk’s central arguments are certainly valid (and – as previously noted – covered in much more detail in Nick Bostrom’s book Superintelligence: Paths, Dangers, Strategies), personally I can’t help thinking that any robotic superintelligence would be more preoccupied with getting away from this paper-thin envelope of breathable atmosphere on the 3rd rock from the Sun and going off to explore the galaxy than engaging in any kind of Terminator-style war of attrition with baseline humanity. My philosophical view is that it is far more likely to view us with the same casual disinterest with which we humans view, say, termites in a termite mound today. Reassuringly there are still plenty of termites left in the world today. And their understanding of humans is probably about the level that ours would be of any posthuman superintelligence. Rest easy. :-/
A First Look Into ConstructionTech
Living in the city of Christchurch, New Zealand where the economy has been going through a 5-year construction boom thanks to both the 2011 earthquake rebuild and NZ’s fast-growing national economy generally, I am constantly amazed at how primitive the construction industry appears to be in its use of information management and automation technologies. Building sites in 2017 still crawl with people in high-vis, banging steel bolts into place with hammers and manually operating diesel-powered machinery. Projects are still managed using 2D paper plans and printed gantt charts.
Compare this to a mid-future vision of construction where the client walks around in a high-resolution VR rendering of their future building, changing walls, pipes, fixtures and fittings at will and immediately being able to see the price / compliance implications. Once they’re happy with the building they’re commissioning, they effectively click “Buy Now” and this sends instructions to a network of 3D-printers and other manufacturers (possibly on the other side of the world) to make up the modular components of the building and place them into containers to be shipped to site. When they arrive, a fleet of autonomous robots promptly take the pieces out of the containers and clip the building together piece by piece. Building consent is automatically filed with the local authorities. (And it goes without saying that when the building is looking a bit tired in 50+ years time, it can be “unclipped” and removed from site and recycled without the risks of micro dust particles filling the air or filling huge amounts of landfill with rubble.)
In our current overinflated house price bubble market of 2017, a residential house or commercial property just doesn’t seem to deliver anything like value for money. Buildings and property are effectively being used as a “reserve currency” rather than for housing itself – and the value of these properties can go down as well as up. Seems to me that construction is one of those asset-heavy industries which is ripe for disruption. And it seems the VCs of the world agree with nearly $1Bn being invested into ConstructionTech over 2015-2017.
Here’s a market map from CBInsights with over 100 ConstructionTech businesses profiled, in spaces including AR/VR, Drones, Risk Management, Marketplaces and Supply Chain Management. There must be thousands more worldwide. Watch this space.
The guys at SpaceCraft who have designed an “open-source” Wikihouse built out of plywood have an alternative approach to transform the economics of the construction sector.
There are 3D-printed houses that take less than 24 hours to construct.
You can even potentially grow a house out of grafted trees or (ew) synthetic meat.
Wealth Inequality – Facts and Figures
This online resource from Inequality.org starkly illustrates the uneven (and getting more so) distribution of income and wealth around the world, particularly in the US. Facts are facts. An essential resource.
Meanwhile Finland is the latest country to do a Basic Income Experiment, this time for the unemployed. Incidentally this month I was introduced to the concept of a Universal Basic Dividend rather than Universal Basic Income: “…the National Dividend would vary depending on the performance of the economy…” because the existing financial system generates consumer prices at a faster rate than it distributes consumer incomes”. Intriguing. Local NZ progressive politician Raf Manji also makes The Case for Economic Rights this month. Something has to be done, right?
The Principles Of Scalability
Philosopher Sam Harris talks to Geoffrey West, author of the recently published Scale: The Universal Laws of Life and Death in Organisms, Cities and Companies , in this podcast “From Cells To Cities”. Massive wide-ranging 2-hour conversation on the similarities between how biological and social systems scale, the significance of fractals and emerging properties in complex “life-form”-like systems like cities. (Starts from around 8 minutes in). Fascinating. Can’t wait to read the book.
A Use For Blockchain?
The self-proclaimed “CryptoValley” Swiss town of Zug announces its implementation of citizen IDs using Ethereum-based Blockchain tech.
Cryptocorrection
Following on from last month’s reading list, the Crypto bubble has deflated a bit in the last few weeks.
Meanwhile this podcast from Tim Ferriss talking to Crypto guru Nick Szabo – the guy who originally coined the term “smart contract” and arguably conceived of Bitcoin itself – is hugely insightful, helps to understand the fundamental (non-speculative) value of cryptocurrencies going forward.
Adoption Chain Risk
I enjoyed this article from prolific blogger and SaaS VC Tom Tunguz, introducing the concept of Adoption Chain Risk: “the extent to which partners will need to adopt your innovation before end consumers have a chance to assess the full value proposition.” This is a problem I’ve seen time and time again for SaaS startups that need to engage with the whole value chain, not just their immediate customer. Adoption Chain Risk – The Importance Of Selling To Everyone In Your Startup’s Supply Chain
(This Week) Coffee Is GOOD For You
Drinking more coffee is associated with a longer life, according to new research. Just as well, eh?
Speaking of Addictions…
“In Silicon Valley, it’s a race to the bottom of your brainstem; where your fear, anxiety, and loneliness reside…” You vs Technology
Stressful
How stress works in the human body to make or break us. Long Aeon essay on how the “subtle flows and toxic hits of stress get under the skin, making and breaking the body and brain over a lifetime”.
The End Of Guessing?
Let’s leave off with this pearl which popped up on my LinkedIn feed last week (ignore the US-centric brandnames…):
The end of guessing:
We don’t guess the weather. We use the weather app.
We don’t guess for food delivery. We use Seamless.
We don’t guess for directions. We use Google maps.
We don’t guess to get a taxi. We use Uber.
We don’t guess for getting the best hotel or flight deal. We use Kayak.
We don’t guess for restaurants. We use Yelp.
We don’t guess for getting a contractor. We use Angie’s list.
We don’t guess for job salary. We use Glassdoor.
We don’t guess with our workout. We use Fitbit.
We don’t guess for getting the best rate. We use Bankrate.
We still guess for healthcare costs.
We still guess if we can pay off our student loans.
We still guess on TV advertising.
The more guessing is in an industry, the more likely it is to be disrupted.