Get in touch
555-555-5555
mymail@mailservice.com

Artificial Intelligence

How does it work and where is it going?



In 1984 I left British Gas to take up my first full-time lecturing post. I had previously worked in electronics, telecommunications, computer maintenance, computer science and also in robotics. I then went on to lecture in microcomputers - microprocessor hardware and software. At the point of leaving British Gas where I had worked in microwave radio, telemetry and microcomputing, I received as a leaving present a book on artificial intelligence as this was something that I had become interested in when previously working as a technician in a university computer science department.


Although my career took a different path subsequently, I maintained an interest in computing and also in robotics. Over the years I read one or two books about artificial intelligence which was moving forward at a fairly rapid pace, but nothing compared to how it has progressed in the last few years, or even the last couple of years.


I am currently reading The Coming Wave, by Mustafa Suleyman and although this is not the reason for me deciding to immerse myself in learning about the current state of artificial intelligence developments, it is in fact a very timely publication and is highly recommended on an international level. This, not least because Mr Suleyman is considered to be an expert in the field having been a co-founder of Deep Mind, which is responsible for bringing the (computerised) game of chess up to and beyond grandmaster level. He is also the creator of pi.ai which is a chatbot produced by his present company and is also considered to be an authority on chatbots and artificial intelligence and its projected effects on all of technological (and biological) progress in the foreseeable future. Whichever way you look at it, Suleman is a technologist, an entrepreneur and most definitely a visionary. Because of his involvement in this field, whatever he has to say on the subject needs to be listened to, as he is not simply an observer, but has contributed massively to state of the art artificial intelligence and is now well placed to be consulted on the effects and implications of AI.



AI is a subject which I have been interested in for a long time, not least because of my technical background. Despite my technical knowledge, experience and commitment to computing and all of its works, I resist new gadgets, apps, signups and linking of devices and so on. But like everyone else, I find the internet a fantastically useful source, along with emails, attachments, Whatsapp, etc.


My initial reaction to artificial intelligence, specifically developments over the last year or so, would be to try and ignore it and certainly not succumb to its obvious temptations. Not only is this an impossibility, as I am sure that you are aware, on the one hand the engine is running and it will be unstoppable and on the other, it will be fully embraced by all those who see its massive potential, in particular, from a personal point of view.


After leaving my job at Warwick University in 1975, I continued my interest in AI and its development. Indeed,I purchased a couple of books over the last few months and purchased the Suleyman book only a month after it was published.


I decided to make this a serious project. Rather than embrace the subject and play with its manifestations (chatbots, image creators and so on), my objective is to write about it as a means of understanding and clarifying this extremely complex artefact and to be able to discuss it with some substantial knowledge.


I do not intend to write a book, there being umpteen of these about and I suspect many people will not be interested in reading about it (but more interested in playing and utilising available AI software for work, etc).


A massive subject and learning curve and also a fascinating intellectual exercise, but also I intend to be ‘on a mission’ to educate wherever there is an opportunity. Before embarking on this I was considering the effects of AI on the unsuspecting masses, including children. And also the government, who in my mind would not necessarily see what is coming and certainly will not be able to control it. This despite the intellectuals and academics at their disposal. I often state, somewhat cynically, that we are not controlled by government (apart from the bureaucrats therein), but to a large extent, by behemoths such as Microsoft, Apple and Google. But to put it another way, we are controlled by technocrats! The internet/Google/Amazon massively affect nearly every citizen, so in a sense we no longer live in a democracy, but a technocracy ….



A ‘Language Learning Model’ typically refers to a type of machine learning model or artificial intelligence system that is designed to understand, process, and generate human language. These models are used for a wide range of natural language processing (NLP) tasks, such as text classification, sentiment analysis, machine translation, chatbots, and more. They are particularly adept at working with unstructured text data.



Transformer models, like BERT, GPT-3 and their variants, have revolutionized the field of NLP (Natural Language Processing). They excel at capturing contextual information in language and are pre-trained on vast corpora of text data, making them highly versatile for various language related tasks.



A chatbot is an artificial intelligence artefact which accepts messages and requests from a user via a keyboard and like Google searches a database of knowledge which it retrieves from its vast memory. Unlike Google, it does not simply produce chunks of information in the form of texts and images, but it is trained to not only interpret words inputted from the keyboard but ‘understands’ and attempts to reply accurately in ‘good English’. This is not a crude a selection of paragraphs or pages, but it would appear that it is conversing in exactly the way that if a message or request was presented.


If you have a ‘conversation’ or dialogue with for example OpenAI’s ChatBot GPT-3 or Inflexion’s pi.ai you should expect to experience a converstion or dialogue as if you were texting a real person.


Whereas the answers you receive may not be perfect and you may realise this, you will almost certainly be communicated with in perfect English.





Share by: