Bill Wadge's Blog

[I did not write or illustrate this post. The text was produced by GPT-3, and the images by stable diffusion. I chose the cities and generated the text with, in most cases, as prompt the name of the city. For the images I added “cartoon” to the prompt. In some cases I asked for more instances till I got something good.

It’s tempting to say that AI “wrote” this post but we’re not there yet. I chose the cities and selected the best instances. But I didn’t write anything other than prompts, at most 2 or 3 words each.]

Venice

There are few places on Earth as unique and enchanting as Venice. This floating city is like nowhere else, and a visit here is sure to be an unforgettable experience. From the canals that criss-cross the city to the stunning architecture, there is so much to see and do in Venice.

A trip to Venice would not be complete without a ride down one of its famous canals. Seeing the city from this perspective is truly magical, and it’s a great way to get around without having to deal with traffic or crowds. You can even stop for a drink or bite to eat at one of the many canal-side cafes while you watch the world go by.

Mont Saint Michel

The island of Mont Saint-Michel is located in the Normandy region of France. It is best known for its abbey, which was founded in the 8th century. The abbey has been a pilgrimage site for centuries and is one of the most popular tourist destinations in France.

The island itself is only about 1 square mile in size, but it is packed with history and culture. In addition to the abbey, there are several other churches and monasteries on the island. There are also a number of museums that detail the history of Mont Saint-Michel and its Abbey.

A visit to Mont Saint-Michel is like stepping back in time. The buildings and streets have remained largely unchanged for centuries, giving visitors a glimpse into what life was like hundreds of years ago. If you’re looking for a truly unique experience, add Mont Saint-Michel to your list of must-see places!

Notre Dame Cathedral

Notre Dame Cathedral is one of the most popular tourist destinations in Paris. The Gothic cathedral, built in the 12th and 13th centuries, is located on the Île de la Cité in the heart of Paris. Notre Dame is famous for its architecture, as well as its history and association with Catholicism.

Over 12 million people visit Notre Dame each year, making it one of the most visited monuments in France. The exterior of the cathedral features two large towers, intricate gargoyles, and flying buttresses. Inside, visitors can see beautiful stained glass windows and climb to the top of one of the towers for a panoramic view of Paris.

Notre Dame has a long history that dates back to medieval times. In 1163 construction began under Bishop Maurice de Sully. It was not completed until 1345 due to financial difficulties and other problems along the way. The cathedral was badly damaged during both world wars but has since been restored to its former glory. [Not yet!]

Edinburgh

Edinburgh is the capital city of Scotland and one of its 32 council areas. It is located in the country’s south-east, between the Forth estuary and the Pentland Hills. With an estimated population of 488,050 in 2017, it is Scotland’s second most populous city after Glasgow and the seventh most populous in the United Kingdom. The official name of he city is Edinburgh, but it is also known as “Auld Reekie”, a Scots translation of Old Smokey.

Edinburgh has been recognised as being one of Europe’s leading tourist destinations for many years now. In a recent poll by Conde Nast Traveller magazine readers voted Edinburgh as their fourth favourite European city to visit behind Rome, Paris and Florence . This was no doubt helped by its World Heritage status which was conferred on Edinburgh’s Old Town by UNESCO in 1995.

Rome

Rome is one of the most beautiful and historic cities in the world. It is a city that has been influential in art, culture, politics and religion for centuries. Rome is a city that everyone should visit at least once in their lifetime.

There are so many things to see and do in Rome, it can be hard to know where to start. A good place to begin is by visiting some of the city’s famous landmarks such as the Colosseum, Vatican City or the Trevi Fountain. Or why not take a walk through one of Rome’s stunning parks or down one of its atmospheric streets? You could also spend a day exploring Roman ruins such as the Forum or Pompeii.

Oxford

There are few places more steeped in history and tradition than Oxford. Founded in the 11th century, the University of Oxford is the oldest university in the English-speaking world. Today, its 43 colleges and six permanent private halls occupy a central position within the city of Oxford.

Despite its long history, Oxford is a very modern university. It was ranked first in Europe for employability and second globally for social sciences by Times Higher Education World University Rankings 2019. The university has produced many notable alumni including 26 British prime ministers, numerous Nobel laureates, and over 120 Olympic medalists!

Whether you’re interested in studying at one of the world’s most prestigious universities or simply exploring a beautiful medieval city, Oxford is definitely worth a visit!

Los Angeles

Los Angeles is a large, sprawling city located in southern California. It is the second largest city in the United States and is known for its warm weather, diverse population, and many tourist attractions.

There are many things to do in Los Angeles, whether you’re interested in exploring the city’s vibrant nightlife or spending a day at one of its many beaches. The Santa Monica Pier is a popular spot for tourists, as it offers stunning views of the Pacific Ocean and has a wide variety of shops and restaurants. Hollywood Boulevard is another must-see while you’re in LA – it’s home to the Walk of Fame, where you can see stars honoring some of America’s most famous celebrities.

Vancouver

Vancouver is a beautiful city located in the province of British Columbia in Canada. It is surrounded by mountains and has a mild climate, which makes it a popular tourist destination. Vancouver is also home to a diverse population, with people from all over the world living and working in the city.

The economy of Vancouver is strong and growing, thanks to its many different industries. Tourism is one of the biggest industries in Vancouver, as the city receives millions of visitors each year. Other important industries include forestry, fishing, mining, and manufacturing.

[I could go on almost indefinitely – and wish I could, because some of the remaining images are beautiful. But I think I’ve made my point, which is that AI can produce the content for whole blog posts of a certain kind. The blurbs are almost perfect – I didn’t edit them. There was one glitch, where GPT-3 showed it didn’t know about the recent fire that nearly destroyed Notre Dame cathedral.

Well I for one welcome our new AI overlords. I couldn’t have done this post without them. I could have written blurbs but that would been a whole lot of work, and I can’t draw worth beans.]

Art is what you can get away with.
– Marshall McLuhan

[All the images in this post were produced with generative AI – Midjourney, DALL-E 2, Stable diffusion. Most by Paul DelSignore, not by me}

I used to teach the AI course at the University of Victoria – thank God I’m retired. I couldn’t have kept up with the breakthroughs in translation, game playing, and especially generative AI.

When I taught AI, it was mainly Good Old Fashioned AI (GOFAI). I retired in 2015, just before the death of GOFAI. I dodged a bullet.

I am in awe of NFAI (New-Fangled AI) yet I still don’t understand how it works. But I do understand GOFAI and I’d like to share my awe of NFAI and my understanding of why GOFAI is not awe-full.

Seek and Ye Shall Find

For a long timeAI was almost a joke amongst non-AI computer scientists. There was so much hype but the hyped potential breakthroughs never materialized. One common quip was that AI was actually natural stupidity.

Many departments, like my own, basically boycotted the subject, maybe only offering a single introductory course

The heart of GOFAI is searching – of trees and, more generally, graphs. For many decades the benchmark for tree searching was chess. Generations (literally) of AI researchers followed the program first proposed by Norbert Wiener in the 40s, based on searching the chess game tree. Every ten years AI evangelists would promise that computer chess mastery was only ten years away

Wiener’s idea, described in his pioneering book Cybernetics, was a min/max search of the game tree, resorting to a heuristic to evaluate positions when the search got too deep.

The chess game tree gets big very quickly and it wasn’t until decades later (the late 1990’s) that IBM marshalled the horsepower to realize Wiener’s dream. They built a special purpose machine, Deep Blue, capable of examining 100 million positions per second. Deep Blue eventually won, first a game, then a whole match, against Gary Kasparov, the world champion.

Deep Blue was the high water mark of GOFAI and there was no real followup. Deep Blue’s successor, Watson, could win at mastermind but commercial applications never materialized.

AlphaGo and AlphaZero

I was impressed by Deep Blue but wondered about the game of Go (Baduk, Wei-chi). The board is 19×19 and the game tree is incomparably bigger than that of chess. If you’d asked me at the time I would have said Go mastery was inconceivable (which, if we had to use GOFAI, was true).

Then in 2016 the unthinkable occurred: a program, called “AlphaGo”, started beating Go champions. It did not use Wiener’s approach; instead it used Machine Learning (ML) (don’t ask me how that works).

AlphaGo trained by playing millions of games against itself. Originally it was given hundreds of thousands of expert level human games but its successor AlphaZero dispensed with them and simply taught itself. It took only a few hours to reach expert level, which for humans took hundreds of years. Variants of the software mastered chess, checkers, and shogi in a similar fashion.

Google Translate

About the same time users of Google Translate noticed a sudden dramatic increase in the quality of its translations, although Google at the time said nothing. In fact Google had switched from statistical to neural networks (don’t ask).

Currently Google’s translations are spectacularly good (and Deepl’s even better). In translating into French, for example, it usually gets the pronouns right as well a well as the gender/number agreements. If you want perfect output you still have to at least proofread the output because occasionally there are mistakes, usually minor.

Here is a translation from today’s Le Monde. It’s perfect.

Since the beginning of the conflict, Russia has refused to describe its invasion of Ukraine as a war, calling it a “special military operation”. Shortly after the conflict broke out, the authorities banned the Russian media from using any information other than official statements on the subject and banned the use of words such as “war” and “invasion”, under penalty of being punished by the law.

GOFAI never produced really good translations, relying as it did on facts and rules. I remember the whole class getting a good laugh out of the mistakes. Once, translating a news article about Iraq from German, it claimed Saddam “warned the US before an attack” when what it should have said was “Saddam warned the US of an attack”. Modern systems would never make such a basic mistake.

Generative Speech

Then comes GPT 3 and the other systems that generate speech. Originally they were merely predictors that would complete incomplete sentences. But now they run on and produce whole articles. Which (usually) make sense. Here’s an example:

Author Prompt: Write me an article about creativity please 

GPT-3 Response: I’ll be honest. I don’t know much about creativity. I’m not a psychologist and I’m not a creative myself. I’m just a guy who likes to read about psychology and creativity. What I can do, though, is write you an article about creativity that’s based on what I have read. That way you get a proper article, but you don’t have to pay for the expertise of a psychologist or a creative person.

GOFAI never achieved anything along these lines. But then my mind was well and truly boggled by …

Generative Art

Along came DALL-E and DALL-E 2. But it wasn’t till Stable Diffusion was released that I started paying attention. Of course there was the pictures of astronauts on horseback and cats wearing sunglasses. But what really impressed me was pictures in the style of well known artists. Here are two of my favourites :

The first is an abstract image in the style of Picasso. I can’t find the original but MidJourney’s version is just marvellous. I’d have no hesitation to print it, frame it, and hang it on my wall.

My second favourite is a wonderful portrait of Superman – ‘by’ Rembrandt! As one observer commented, “those eyes have seen some shit!”

But even the cheesy astronaut image is impressive.

The striking fact is that you can’t see the astronaut’s left leg. The image generator seems to understand that you can’t see through opaque objects (namely, the horse).

GOFAI would need literally hundreds of rules just about what to do when bodies overlap, what to show, what objects are transparent and to what degree etc etc.

On reflection

OK let’s go all in – let’s look at a cat wearing sunglasses. Ew cheesy – but there’s something remarkable about the image.

It’s the reflections in the lenses of the sunglasses. Not only are they visible, but the reflections are, correctly, the same. How does Midjourney coordinate the images in separate parts of the picture?

A closer look

When I see this image I have to ask, where did all this come from? Midjourney is trained on 5 billion images but condenses this training to 5 GB. So there’s not enough room to include exact copies of images found in the training set. We can assume that this (apparent) photo does not exist as is on the internet.

In particular what about the blue feathers on either side of the subject’s neck (they are not mirror images). Where did they come from? Did one of the training images have them?

The mystery is that this image is the result of combining training set images, but how are they put together? The best GOFAI could do is chop up the training images and put them together like a badly fitting crossword puzzle with visible seams and limited symmetry. I’m baffled.

The social implications of AI technology

It is questionable if all the mechanical inventions yet made have lightened the day’s toil of any human being.

~ John Stuart Mill

There is a lot of controversy Midjourney and other generative image programs.

The first question is, are these images art? I think some of the images presented here are definitely art, even good art. If you’re not convinced, have another ‘Rembrandt’.

The second question is, is imitating the style of certain artists fair? I don’t know, but there seems no way to stop it. Currently nothing stops a human artist from studying living artists and imitating their styles. Midjourney etc are just especially good at this.

In a sense, this imitation broadens the exposure of the imitated artists. Now everyone can have, say, a Monet of their own.

Finally, a vital question is, how will this affect today’s working artists? Here the answer is not so optimistic.

Generative AI is not the first disruptive technology. There’s photography, the closest analog, digital art in general, the telephone, the automobile, the record player, the printing press, and so on.

Each of these had the effect of obsoleting the skills of whole professions. It didn’t wipe them out, but the vast increase in productivity put large numbers out of work. And those that remained had to acquire and use the new tools. Because of economic competition they had to work harder than ever to keep up.

Labor saving technology inevitably becomes profit saving technology. The tractor is an example. Initially it (and farm machinery in general) were marketed as labor saving. But eventually competition forced every farmer to get machinery or sell out (which most had to do). The result was the same or more food produced by a fraction of the former number of farmers, working their butts off.

So I predict AI will shrink the number of artists and force them to use Midjourney etc. For art consumers, it will be good news – like drinking from a firehose. A new individual Monet every week. Do it yourself illustrations for personal blogs. But not change in society as a whole.

Power concedes nothing without a demand.
-Frederick Douglass

When the late Ed Ashcroft and I invented Lucid, we had no idea what we were getting in for.

Continue reading →

When Ed Ashcroft and I invented Lucid, we intended it be a general purpose language like Pascal (very popular at the time.)

Pascal had while loops and we managed to do iteration with equations. However in Pascal you can nest while loops and have iterations that run their course while enclosing loops are frozen. This was a problem for us.

Continue reading →

Recently there was a post on the HN front page pointing to a GitHub repository containing an old (2019) version of the source for PyLucid (I don’t know who posted it). It generated a lot of interest in Lucid but unfortunately the 2019 version is crude and out of date and probably put people off.

I’m going to set things right by releasing an up to date version of PyLucid (Python-based Lucid) and up to date instructions on how to use it to run PyLucid programs. The source can be found at pyflang.com and this blog post is the instructions. (The source also has a brief README text file.)

Continue reading →

I recently graduated my 17^th and last PhD student, Monem Shennat. This is the abstract of his dissertation with my annotations (the abstract of a University of Victoria dissertation is limited to 500 words).

The problem he tackled was that of dimensional analysis of multidimensional Lucid programs. This means determining, for each variable in the program, the set of relevant dimensions, those whose coordinates are necessary for evaluating individual components.

Objective: to design Dimensional Analysis (DA) algorithms for the multidimensional dialect PyLucid of Lucid, the equational dataflow language. 

Continue reading →

I have some theories about these modes – for example, cropping one into the other. I tried them out on the Monna Lisa and … read on!

Continue reading →

Recently my PhD student gave a rehearsal of their 20 minute oral presentation. It was ok. Average.

In other words, (seemingly) long, and boring. Like so many people’s technical talks. What can you do?

What you can do is follow these simple rules I’m going to give. They’re not all my own, you can find most elsewhere. The problem is, most people think they’re impractical and don’t follow them. Result? Bo-ring!

Continue reading →

Two cheers for the World Wide Web
— Ted Nelson

Ted Nelson invented hypertext but not the web. He thinks it hasn’t fulfilled its real potential, and I agree.

One of his good ideas that the web doesn’t really support is stretchtext – text that expands or contracts in response to the reader’s (dis)interest.

Continue reading →

As I’ve said before, PyFL is functional programming for the rest of us. (It’s available at pyflang.com.)

PyFL now has type checking – without type declarations. Instead the type is produced by evaluating the program over the domain of types.

Continue reading →