This year's Nobel Prize drama, Human workers in Asia are still cheaper than robots
Google, all the things! and Greatest Events of WWII in Colour
I hope you had a good week, My gift to you today is not to cover anything about Disney+, but I can’t promise that I won’t cover it in future issues. If you really can’t wait, I recommend that you listen to this podcast episode from Recode Media.
This year's Nobel Prize drama
Do you have a very smart friend that never gets a haircut except for once a year by the end of September? If yes, I’m afraid to tell you that your friend is an aspiring Nobel prize winner.
Every year around that same time, the new Nobel prize winners get announced to the world, and according to the old tradition, at least one of the prize winners should be the subject of an internet fight.
This year’s internet fight was about 3 of those winners: A. Banerjee, E. Duflo, and M. Kremer won the Nobel Prize in Economic Sciences for their experimental approach to alleviating global poverty. The 3 economists have been strong contributors to the increasing adoption of RCTs (Randomized Controlled Trials, A/B tests) to tackle poverty problems.
The award divided the economists on the internet into 2 groups:
The first, happy for the winners and excited about the increasing use of RCTs in development economics.
The second, angry about that choice and undermine the value of RCTs. Their argument against RCTs is that they are hard to implement accurately and that they drive solutions that are hard to replicate under other conditions. This group is instead in favor of tackling the harder problem that will have a bigger impact, which is developing economic models that can help solve the root cause of poverty.
Does that sound familiar?
The whole thing reminds me of the same debates Design & Product people get involved in about the use of RCTs (A/B testing) when building products.
Some groups usually blame RCTs for being focused only on solving measurable problems and for being short term focused. While this is not always true (it depends on how you use the tool), this group’s recommended approach is to go after big daring projects that are high in risk, but higher in impact.
I don’t personally believe that one approach is always better than the other, but both approaches should get used whenever and wherever it makes sense. Actually, both approaches can be used in parallel with each other.
In the context of tackling poverty-related problems, there is nothing wrong with running small and focused RCTs that try to find effective solutions for specific problems like, how to get more babies vaccinated? how to get more children to not drop out of school? how to get more poor households to benefit from government-sponsored programs?
While answering one of these questions for the residents of one city in a poor country is not going to solve the root causes of their poverty, still, this answer would help introduce a small incremental improvement to the quality of their lives while they wait for the big high impact solutions to be discovered.
If you have some time to learn more about the use of RCTs to reduce poverty, watch the following talk by Esther Duflo (one of the winners). She gives some examples, and honestly they look very similar to RCTs I run with my teams at work but in a different context. I’m happy to share my learnings with others who work in the field of development.
Google, all the things!
This week, Ben Thompson wrote three times about Google. The first 2 times were about their travel & shopping ad products and the third was about their newest payment product. (Everyone is doing payments now)
Add to that, some news leaked that they have built a search engine for medical records that already has access to the medical records of millions of US citizens. Also, the week before they acquired Fitbit for $2.1 Billion (just 1.7% of Google’s cash reserve). The Fitbit deal gives them access to the health data of 28 million active Fitbit users, which seems to be like a good amount of data to start training some new health-related machine learning models.
While there is some chance that Google is going to use this health data to advertise consumers something sooner or later, most probably the first thing to happen is that Google will start utilizing consumer’s search queries and other information they know about them to enrich the medical records they have, and based on that deliver better health predictions about consumers to their doctors and to the health insurance companies.
I really think that Facebook is getting too much attention from everyone just because of their impact on politics, but the real big unstoppable machine is Google.
Human workers in Asia are still cheaper than robots
If you thought that the robots are about to take over all the jobs in factories, it seems like you’re wrong, at least about some of the robots. Adidas announced this week that they are going to shut down their brand new robot factories located in Germany and the USA.
The new plan is to move the production back to Asia where Adidas indirectly employes 1 million workers through its contract factories. While this is not good news for the American or German factory workers, it’s amazing news for the Asian workers that would have struggled if the robots were ready to take over their jobs.
Considering the rapid development in the robotics industry, I would not be surprised if Adidas decides to give the robots factory another shot in 5 years from now.
While robots are not that efficient when it comes to making shoes, they are still very capable of handling a lot of other tasks.
Oxford Economics shared recently its analysis and predictions about the impact of automation on factory workers. The big prediction they have is that by 2030, around 20 million factory jobs will disappear across the world, but they expect that each new robot will increase the economy’s productivity and as a side effect will increase the demand for new workers in the job sectors that can’t be automated yet.
Considering that the average career for workers is 45 years, this means that the low end of workers (the ones with the highest risk of losing their job to a robot) will be changing careers 2 to 4 times. One time for every new significant improvement to the robotics technology, with the assumption that the robots are not going to accelerate their learning speed over time.
What’s the take away for the young people entering the job market for the first time? look for jobs that are hard to automate. Most probably the ones that are very unstructured, and require a lot of emotions, and human connection.
While I don’t sell ads on this newsletter, I want to use this section to advertise good causes. This time I want to persuade you to consider using Kiva to lend money directly to people in need of microloans that can help them get clean water, buy a new cow, or start some small project to feed the family.
This is not a donation, your money will come back to you in installments over a 1 to 2 years schedule. If you have read the previous issue of my newsletter, maybe you remember the article I shared about the negative interest rates. Maybe this is a better way for you to stash a few of your millions without suffering from the negative interest rates. [This is not financial advice, I’m just throwing ideas].
Greatest Events of WWII in Colour
What am I watching these days, you ask? I’m watching this documentary about WWII on Netflix. As an Egyptian, I didn’t learn much at school about the wars that happened outside Egypt and the Middle East. Sorry, we had our own long list of wars that filled all the time during our history classes.
After moving to Europe, I thought that it might be a good idea to start learning about these other wars I didn’t know much about. Luckily, Netflix and youtube have a rich library of war documentaries that can keep me busy for years.
That’s it for this week. If you enjoyed this issue of my newsletter, forward it to your friends, and ask them to subscribe to the newsletter, and follow me on twitter (@shreef). Also, send me your feedback.
Have a good weekend!