Today's blog is about AI. I touch on the impact of AI, how we can impact the future progression of AI, give a brief introduction to the history and science behind AI and I finish talking about some upcoming content I am planning with a friend. I say planning, I mean releasing, because the content has been created.

AI has been developing incredibly quickly, even if it doesn't always feel that way. A lot of people have had exposure to large language models such as Chat GPT and Gemini, and image generators like DALL-E or DreamStudio, but AI is used for far more than just that. This article on Wikipedia is long and clunky, but it shows many of the areas and fields that AI is used in. You can check the long list here.

I love thinking and conversing about AI, as well as using it. It is a hugely important topic and there are many, many implications. A lot positive, others negative and a multitude that we won't have considered yet. AI has had some incredibly positive outcomes regarding medicine, science, computing, supply chains, art creation, productivity and a whole host of other things.

Does that mean we should ignore the negatives, of course it doesn't. For every positive implication for AI, the equal and opposite negative use is possible. And knowing humanity, likely being played with, if not being used. As more and more AI modelling becomes open source, this risk is only ever growing. I don't think the answer is limiting use, or who can use AI.

I think the answers lie in our questions and our use of AI. I think that by playing with AI (using it to help with our day to day tasks) we naturally test the boundaries of what it is capable of and what is acceptable. One question is should AI art be watermarked? It's something I have argued for in the past, now I am not so sure (and I could write a lot just on that!). We need to experiment, push the boundaries, question and get questioned by our peers. That's the way we work out the rules.

There are three ways that anyone can react to AI, and I think there is only one reaction that means you can have an impact on the future of AI. We can be fearful and watch, we can watch with intrigue and an open perspective, or we can get stuck in and embrace AI. Embracing AI doesn't mean celebrating every use of AI, or that you have to be happy about the job impact, but it does mean using it. I cannot advocate for that enough. We need to use AI.

We need to use AI to impact it's growth and where it goes next. AI learns from humans. You could look at it like we are it's parents and it will mimic and copy behaviour we exhibit, as well as wanting to please us with the results it populates. And think about that last point. AI large language models will want to please us, it's what they were designed for. To extrapolate information from patterns and provide answers/solutions to our queries. Answers and solutions that we want.

It wants this because they have been built that way. They were built that way because they drive revenue to the companies that build and house them. Those companies want you to continue using their models so you continue to drive revenue their way. If any given AI model is not giving you what you want, you will go and use a model that does give you what you want!

I wonder if I should have started here, with a brief history of A, but we are here now. And this is where I am sharing it. AI was conceptualised during 'The Golden Age of Science Fiction' in the 1930's and 1940's. And then humanity started working on AI in the mid part of the 20th Century. Alan Turing wrote a paper call Computing Machinery and Intelligence in 1950, where his first question is "Can machines think?". I've started reading this paper but not finished it, the rest of it is now on my reading list!

Computing limits meant even exploring AI wasn't possible until the mid 1950's, until more computers had memory capacity. Logic Theorist was a programme written in 1956 and is often considered the first AI program. Until the mid 1970's AI improved drastically and quickly, mainly due to computers storing more information and becoming faster. Development then stalled for a few years because we had to wait for more computing power.

In the 1980's there was a funding boost, deep learning techniques were developed and decision making processes were improved drastically. In the 1990's improvements continued. Deep Blue beat Gary Kasparov at chess in 1997, the same year speech recognition was added to Windows. Kismet was a robot whose life began in 1997, a robot the could read, respond to and display human emotions by 2001.

Development has accelerated into what we know and use today. The areas of AI are Reactive Machines. Limited Memory Machines, Theory of Mind and Self Awareness. Reactive Machines and Limited Memory Machines exist right now, we use them all the time. Machine Learning and Deep Learning are systems behind the way that AI works.

An article on the BBC from 25th May 2017 when Google's DeepMind AlphaGo AI defeated the then world number 1, and current number 2, Go player, Ke Jie. Read more here.

I don't want to get too stuck on history and science but I think it's important to build a basic understanding of the history and science behind any subject I want to explore in more detail, or talk about with any authority. I want to round this part off with an explanation about the differences between Machine Learning and Deep Learning. Both huge parts of Artificial Intelligence.

Machine Learning is programmes/systems that learn and adapt automatically from experience and with minimal human interference. Deep Learning is a machine learning technique that layers algorithms to mimic a neural network. Deep Learning is part of Machine Learning, which in turn, is part of Artificial Intelligence.

Some examples of Machine Learning are facial recognition, email automation & spam filtering, predictive analysis, the list does go on. Deep Learning is used for speech recognition, image recognition, finding patterns in a dataset, object classification in photographs, character text generation, self-driving cars, among other things.

Apparently today is more article than blog, but that's OK. I still enjoyed writing it. One of the topics I wanted to discuss but haven't, is how to use AI effectively. Very briefly, it's all in the prompt. All of it. It's how you write your prompts. Be clear and concise. AI cannot make what you want unless it knows what you want. I use AI for all sorts of things, I use it to generate promotional material, as feedback on documents, to bounce idea's off of, to make images, I've even made a song, as well as many other things. We all use AI in our day to day lives even if we don't realise it.

I can't quite say yet, but I am hopeful I will be able to spill the beans in my next blog, maybe even share the content myself and a friend have been working on. It's been really exciting playing with a twist on an existing format. We both think it is quite novel and we have enjoyed the process immensely. We nailed down the content we wanted and since then we have been working on all the other things you need for a new 'brand'.

We need to sort out simple things like brand imagery (logo, what font's we are using, promo posters etc.), platforms to put the content on, promotional material and much more. Saying that feels like we have barely started but I know we have done far more than barely start. I'm looking forward to sharing more with you, not yet, but soooooon!

I would love for more people to experiment with AI. Whether that is having a chat with Chat GPT or using a platform like Night Café to generate images. Experiment and explore, and share your learns with others. Articles v blogs is an interesting one for me, and I want to release more articles whilst maintaining my blog. A reminder that my blog is now released every 2 weeks. The next one will be on the 27th May.

If you want to explore my thoughts on AI from around a year ago, you can here. Today's issue has 10 collectible cover images. Once they are gone, they are gone. You can get yours for 5 $Matic including the platform fee.

As ever, thank you for reading and for your time. I hope you have a great couple of weeks and I will catch you soon.