Sep 2, 2023 4 min read

Weekend Briefing No. 14

I'll see you on the flip side.

Good Saturday morning and Happy Labor Day weekend friends! The summer is winding down and I'm going to squeeze out the last bit of fun this weekend before it's over. Enjoy this short briefing and get outside with your loved ones.

Interesting data points

The United States generated 772.2 billion kWh of electricity from nuclear energy in 2022; China is in second place with 395.4 billion kWh
In 1963 there were 96 births per 1000 women aged 15 to 19 globally, in 2021 that number dropped to 43
The median age in the United States is 38.1 years

France has too much wine

This is terrible news but France has too much wine and it's destroying nearly 80 million gallons of it in a bid to keep prices high.

Making wine is getting more expensive due in part to recent world events, and people are drinking less of it. That has left some producers with a surplus that they cannot price high enough to make a profit. Now, some of France’s most famous wine-producing regions, like Bordeaux, are struggling. - Washington Post

The article goes on to say that peak wine consumption per person happened in 1926 when the average French person consumed 136 liters (36 gallons) of wine per year. That number is closer to 40 liters (10.5 gallons) today.

This bothers me a lot, the waste of such a bountiful harvest.

NixOS

This week I came across an old, but new to me, distribution of Linux called NixOS. I discovered it by stumbling across this YouTube video. What caught my eye on this particular Linux distribution is its package manager and the ability to do easy upgrades and rollbacks if needed.

NixOS uses a declarative configuration (i.e. like YAML) to run, upgrade, download, and manage itself. This makes NixOS a robust system for easy reproducibility. You can build 'dev' machines and then bundle it all up and have it run on a different machine with no problems.

I'm going to check this out for sure.

Are we all programmers now

My old college professor sent me this article to get my thoughts on it. The article is "all in" with Generative AI (Gen AI) and discusses how "citizen developers" are going to be using all these Gen AI tools and build all sorts of amazing applications. I vaguely remember another "citizen" type of developer too, the citizen data scientist.

ING bank, based in Amsterdam and operating in 40 countries, used a similar process when it needed to develop more machine learning (ML) models to put into production. Facing a lack of professional data-science talent in many of the countries in which it operates, ING began to explore citizen data-science capabilities. The bank, where one of the authors of this article (Kerem) was recently the global chief analytics officer, is working to supply citizen developers with technical expertise and to identify the use cases that are possible with automated ML. There is no doubt that ING’s employees can create some ML models—to predict, for example, the probability that customers will click on an app message or respond to an email campaign. That can free data-science professionals from doing simple and repetitive data-management and analytics tasks. But aspiring citizen developers at the bank need proper training and hands-on experience to be successful. The ML models built by ING’s employees also need to be free of strict regulatory requirements, though some documentation may be required for them. Finally, IT must still deploy and manage the tools and platforms needed to support automated ML development and use. Those are just a few of the immediate issues ING is addressing as it constructs a policy for citizen development.

Let's be honest here, the citizen data scientist spiel didn't exactly work out. Did some smart citizen data scientist people become real data scientists and got job offers? You bet! However, how many companies would entrust their competitive advantage (and millions of dollars) to someone who can slap together Python code, build a classification model, and rave at it's 99.9% accuracy? Not many, I'm sure.

I see this happening here again as Gen AI will make some tools more approachable for mainstream users, but are you really going to deploy what a Gen AI model creates from a prompt without validating it? No, you won't! And what are the requirements for a good validation? Having people with experience and the right skills to make sure it makes sense.

I believe that any Gen AI product or service will, in the end turn into a big support function. Will Gen AI code completion software replace software engineers? Yes, it will replace some but it will the rest of the software industry leaner and more efficient. Gen AI will compress the time it takes to make things but you will still need to check that it makes sense. The old adage of "garbage in, garbage out" remains here.

End Notes

There's not much to write this week except the disbelief that observed summer season is over this Labor Day weekend. There will be barbecues, fun in the sun, and maybe one more trip to the lake or beach. Whatever your plans may be, I want to wish you a safe holiday filled with wonderful memories. I'll see you on the flip side.