Humanist Association of Toronto (HAT) Meetup Group

The Humanist Association of Toronto
http://www.humanisttoronto.ca/

Every Saturday we meet on Zoom to discuss a topic decided upon the previous week. These are topics of humanist interest, from a humanist perspective.

The topic of the discussion will be decided in a prior meeting, usually two weeks in advance. This week’s topic is: Artificial Intelligence and the Arts by Sonia Brock

| AI & The Arts |
| ------------- |
| |
| I am taking a Course at https://education.gale.com/l-tplmain/. My course in called Introduction to Artificial Intelligence. GALE has many College level course on many subjects which are free with your Library card.


There are various forms of artificial intelligence (AI). Business, the Arts and Industry are being transformed or disrupted by AI implementations such as natural language processing, forecasting, and robotics. Especially in the Arts, there are ethical factors in AI deployment.

The course Instructor, David Iseminger, is an author and technology veteran with expertise in computing, networking, wireless and cloud technologies, data and analytics, artificial intelligence, and blockchain. While with Microsoft, David worked on early versions of Windows and its core networking infrastructure, transmission protocols, security, data visualizations, and multiple emerging cloud technologies.

I’m in so far over my head here that I’m treading water and near drowning. Aside and apart from that, I do have personal experience with AI both as an amateur writer and in the use of AI to create illustrations for short stories I have written.

Graphic Arts

One story I wrote was about a Housebot, a robot for cleaning house which was also a mechanism for spying. It’s not a great story but why I am mentioning it is to illustrate the use of AI as a co-creator. https://soniabrock.com/HB-21.html


First of all I made the illustration using Illustration by Image Creator from Microsoft Bing

I told Bing I wanted a small image of a Robot shaped like a garbage can with arms and fingers and a small red knob on top. Bing gave me several examples and gradually we, together narrowed down the appearance until we got to one I could use. It’s not hard to do. You just have to be very specific telling the AI image creator exactly what you want. Seems fairly harmless, doesn’t it. Hold that thought. Let’s look at the ramifications for the Arts.

Any number of graphics programs will let you manipulate images.Amateur ‘artists’ are using AI image manipulation to put the faces of celebrities and persons known to them on bodies stolen from pornography. That’s the lower side of this new technology.Faces you see online are beautified so the real owner is made glamorous. That’s fairly harmless but you might get a shock if someone you connect with on an online dating service is less attractive in person.Fake news videos and images show us things that never really happened, That’s a bit more serious.Hollywood loves CGI (Computer-generated imagery) Saves a lot of money and distorts reality. It’s a lot easier to send a computer generated car over a cliff and you save a fortune in props and stuntmen.Dead actors can be brought back to screen life with a suitable fee paid to the descendents.

Acting

I could go on for a while here but let’s go on to voice acting. I used to do some Podcasting and it was a lot of work writing scripts and getting the vocals just right. Nowadays I can write a script and get an AI program to read it for me. Here’s the link to a free online voice generator. <https://voicegenerator.io/>

That’s kind of fun. Not so much fun if you’re a voice actor and https://www.audible.ca/ is using AI to read a book to add to their list of literary items instead of paying a real live actor.

AI can replicate the human voice and make a politician or a doctor or anybody say whatever they want.

Writing:

Now, let’s start small and talk about auto-correct.

Microsoft Word and other writing programs now come with auto-correct installed.

It can correct your spelling and suggest words, sometimes with humorous results. Punctuation is added. It’s very handy but you need to keep a critical eye on it.

I am an amateur writer. Sometimes I can’t untangle a sentence to make it clearer. The temptation is great to pop it into ChatGPT and ask the AI to rewrite it, which it does a nice job of in a bland sort of way or not as in the following example where I used ChatGPT and told it to:

Rewrite

"More seriously, I got a granny phone call from someone who purported to be my daughter. AI had duplicated her voice and she needed money right away due to an arrest from a vehicle accident involving alcohol. My daughter doesn’t drive or drink. She’s a Buddhist nun."

in the style of Edgar Rice Burroughs.

The Ai wrote:

More gravely still, there came to me a voice borne across the mysterious ether—a voice I knew as my daughter’s own. Yet it was but a cunning phantom, wrought by some strange artifice of this new age of sorcery we call artificial intelligence. With pleading urgency it declared that she lay in peril, ensnared by the law after a drunken calamity upon the highway. But I, who know her heart, knew it could not be. She touches neither wine nor wheel, and walks the world as a sworn disciple of the Buddha. Thus I discerned the trickery, though the echo of her voice chilled me to the marrow.

Thus, you see, AI can write some really purple prose!

That’s amusing but it’s not so amusing when a real magazine received hundreds and hundreds of AI written Science Fiction stories which swamped their editorial staff. None were publishable. The magazine had to stop receiving submissions for a time and legitimate writers were thus shut out along with the AI fakes.

Newspapers have published articles written by Artificial Intelligence under the direction of a lazy reporter and these articles were, in part, fantasy - not based in real time events and with false quotes and references. AI doesn’t know the truth from false information. It just scrapes available references from the Internet and some of these references come from crackpot conspiracy theorists. Elon Musk has been modifying his GROK AI tool to be more conservative in it’s replies, for instance.

There has been court action in this regard, OpenAI a 63.92 Billion dollar AI model developer and others devour copyrighted fiction and non fiction books and newspapers, such as the NY Times, to base their Large Language Models on.

“Large language models (LLMs) are advanced AI systems designed to understand and generate human language by analyzing vast amounts of text data. They use deep learning techniques to predict and create coherent text based on the context provided.”

That’s how the former rewrite example was developed. It ‘read’ all of the Tarzan, etc. books by Edgar Rice Burroughs and copied his style based on his common use or words and phrases to produce the fake version of the quote I requested.

AI isn’t just taking from literature. It also takes from you the audience of the various arts.

If you have a smart TV, remember it is observing you and your likes and dislikes. Every online store is recording what you buy and from where. Every social media platform whether it be Facebook, Instagram, Whatsapp or YouTube is doing the same. Sure, it’s to find out what you like and sell you more but, more than that, you have been quantified for the purposes of control.

Dystopia or Utopia? We are about to find out. We are living in a controlled environment.

Footnote from my online Artificial Intelligence course:

#The principal constraint in robotic implementations is the need for a controlled environment. Robots use a set of rules, both to carry out their programmed tasks and to interact with their surroundings. Since these rules are rather limited (historically by restrictions such as memory capacity and today by the sheer scope of programming them), a robot needs to perform in a controlled environment.#

|

Meet our diverse group, trade perspectives in a free and open forum and learn from others as they learn from you!

BTW: don't be concerned if there are not many RSVP’s. Many HAT members attend regularly but don’t sign up on Meetup. Our online meetings have been very popular with 20-30 attendees.

NOTE: The HAT Forum adheres strictly to the City of Toronto Policy on Non-Discrimination (http://www.the519.org/public/content/policy-files/The519SpaceUsePolicy.pdf)

Our Website (http://www.humanisttoronto.ca/)