Three news items caught my eye this morning.

Wine to be sold in 1-pint bottles

https://www.bbc.co.uk/news/uk-politics-67795075
27 December 2023

Starting in 2026, the UK government’s Department of Business and Trade has approved the sale of alcoholic beverages, including wine, in 1-pint bottles. Why include wine? A standard bottle of wine contains 750mL, which breaks down nicely to a small (single) measure of 125mL/glass, a medium measure of 175mL/glass and a large (double measure) of 250mL/glass. Those among us who are concerned about how many units of alcohol we are consuming as we drink wine know that a 750mL bottle of white wine with an Alcohol By Volume (ABV) content of, say, 11% contains (0.11 x 750)/10 = 8.25 units of alcohol. Hence, two small or one large glass equals just over three units and puts the wine consumer borderline to the drink-driving limit. Drinking the same quantities of red wine with, say, an ABV of 13.5% will take you over the 3-unit drink-driving limit.

Now consider the 1-pint bottle size. One pint equates to 568mL. What will the mL contents of a small, medium, and large serving of wine be now? 568 is a messy figure. If we stay with the ‘six small glasses from one bottle’ assumption, a small glass of wine will now contain 95mL; a large glass 190mL. A bottle of 11% ABV white wine will contain ((0.11 x 568mL)/10) = 6.25 units of alcohol. ‘Okay, that’s half a bottle and then no more,’ you say, but how will publican landlords deal with this? Will they serve wine in half-pint (284mL) tumblers? I hope not! Or will they have an alternative set of wine measures for dispensing 125mL from a 750mL bottle or 95mL from a 568mL bottle? Or will they throw up their hands in horror and say they will only buy and serve wine in 750mL bottles?

It’s crazy, confusing and unnecessary to bottle wine in 568mL bottles. It serves no purpose, solves no problem, and will almost certainly lead to over-charging for glasses of wine taken from a 1-pint bottle relative to those taken from the well-known 750mL bottle.

Driverless cars by 2026

https://www.bbc.co.uk/news/technology-67825617
27 December 2023

‘Driverless cars could be on some UK roads by the end of 2026, the transport secretary has told the BBC. Mark Harper also said he expected to see the owners of those vehicles being able to travel without having to watch where they’re going by the end of that year.’

So starts this article based on comments made by Rt. Hon Mark Harper MP, the UK government’s Secretary of State for Transport, a position he has held since 25 October 2022. I applaud visionaries. I welcome innovation. I support entrepreneurship. But, I strongly suggest that Mark Harper reads a recent article in the IEEE flagship magazine, Spectrum, written by Mary L ‘Missy’ Cummings, a professor at Duke University and Senior Adviser to America’s National Highway Traffic Safety Administration (NHTSA).

In her July 2023 article, Prof. Cummings summarises several areas of concern regarding the development of driverless cars and their support systems based on large language model AI technology similar to ChatGPT. She discusses:

– the effect of coding errors on large software systems. Such software systems are notorious for containing bugs. Microsoft updates my Windows 11 Operating System every other week, so it seems;

– the inability of an AI-based system to predict and correctly respond to a variety of ever-changing scenarios involving moving and stationary people, moving and stationary vehicles, buildings of various sizes, constantly-changing traffic-control systems, climatic conditions, the time of day or night, and many other independent environmental factors;

– the problem of phantom braking, i.e., when a vehicle suddenly stops for no apparent reason, such as detecting a shadow that is interpreted as another vehicle, for example. (It’s happened in San Francisco);

– the lack of non-partisan regulation, preferably at government level, and the lack of knowledge about large language models by those who might become involved in the requirements of regulation;

– the constant need to retrain or otherwise update the prediction model as factors change e.g., new or unusual vehicle profiles, new roadworks, temporarily shut-off roads, etc.

Professor Cummings has written a stimulating article that Tesla’s CEO, Elon Musk, has criticised, and for which she has received death threats from supporters of the driverless car industry, but, for me, her comments, criticisms and warnings ring very true. I would make her article mandatory reading for all those in Mark Harper’s department working on the new Automated Vehicles Bill.

Jet fuel from human faeces

https://www.bbc.co.uk/news/uk-england-gloucestershire-67771640
27 December 2023

I’ve left the best till last. The title of this article reads:

Firm develops jet fuel made entirely from human poo

Yes folks, it’s true. Firefly Green Fuels, a company in Gloucestershire, has worked out how to convert solid human waste into a grade of kerosene almost indistinguishable from aviation fossil fuel kerosene and eminently suitable to power aircraft. As James Hygate, Firefly Green Fuel’s CEO, said: “We wanted to find a really low-value feedstock that was highly abundant. And, of course, poo is abundant.” And added, “It has no fossil carbon, it’s a fossil-free fuel”.

Human faeces (I prefer this term over the BBC’s euphemistic ‘poo’) is truly abundant and, if the conversion process can be scaled up to industrial levels, could mean that, in the future, we will be transported by aircraft powered by our very own waste. I suspect an outpouring of many humorous expressions will now occur but, for me, the immediate one is the reply to the usual question when you travel somewhere and finally meet the person greeting you at the end of the journey:

How was your journey on Poo-Power Plane’s Airbus?
Crappy, in more ways than one!

(^_^)