Posts

Author: Dave Chapman, Board Member, DDCE

12 June 2023

As a recent retiree and having a little more time to take the news than when I ran an SME for over 25 years, the recent upsurge in artificial intelligence(AI) caught my attention. Today I read that more and more ‘senior leaders’ from both the computing, business and political spectrums are all calling for urgent regulation of AI as there appear to be growing fears that the robots will eventually eliminate humans.

I was in my late teens when I witnessed the birth of computers, mobile telecommunications, and the internet. Being so new it seemed too futuristic to imagine having a device on my wrist that would be a powerful personal computer and able to share live pictures with anyone in the world, instantly and in real time – let alone make and receive phone calls. That kind of stuff only happened in the movies. The early computers ‘spoke’ a different language that you needed to be a ‘boffin’ to understand and were the secret realm of the ‘geek’. Computers could think a million times faster than a human could, could retrieve complex information in the blink of an eye.
This un-known and un-proven technology was however scary to many and there were fears that computers would eventually take peoples’ jobs. Some of this was targeted at welding robots and automated assembly machines in car manufacturing plants.

Due to the vision of a handful of exceptionally talented people, much of the complexity of use was removed and the introduction of software like Microsoft Windows, Apple’s iPhone, Google, and electronic mail transformed computing for both industrial and personal use and with the subsequent development of applications (apps) computers and computing became the norm relatively quickly. Bringing social media into the mix from an early version of Friends Reunited that re-connected old school friends, then Facebook, twitter, Instagram and the like, the world became a very small place and became the steppingstone to an explosion in the use of social media by a younger generation. Add computer games – everything from chess to motor racing to shoot-’em-up war games and the younger generation were fully on board. Some 50 years on it is difficult to image in a world without computers and the modern generation find it near impossible to imagine how we managed without a mobile phone, the internet, and social media.

Did computers take people’s jobs? The simple answer is yes. But they also created new jobs. They helped improve efficiency, and increased the speed at which business could be done, emails being hugely influential in this. I think back to the ‘old days’ of UK car manufacturing, where you hoped that you didn’t buy a Friday afternoon or Monday morning car as these were deemed poorer quality as assembly workers looked forward to the weekend or suffered on Monday because of it. Software facilitated endless possibilities such as desktop publishing that allowed millions of small businesses to produce in-house literature, for example. In my SME we took advantage of new technology that applied to our business as quickly as we could, understanding the cost to time benefit it brings. Early fax machines, scanners and printers come to mind, and of course, mobile phones including a car phone with a briefcase sized battery that weighed as much as a concrete block, took hours to charge and produced 10 minutes of call time!

Computers were certainly the second industrial revolution and it has certainly been interesting to see it from the start to where it is today, although it was impossible to foresee the depth and enormity of its transformational global impact it would have back in the early 1970’s

I wasn’t aware that AI has been developing from the earliest days of computing – its current huge publicity came as a relative surprise to me. I know robotics have been advancing and have seen how amazing the latest Boston Dynamics robots are, and how surgeons use robotic surgery from watching the raft of current medical programmes on TV. Somewhat alarmingly though, I have also seen the ‘soldier robots’ on youTube.

I hear today that the music industry is also calling for urgent regulation because AI can duplicate a singer’s voice and create a new song in minutes, and release it, denying the true artist its royalties. Recently schools were struggling with AI produced homework submissions that can even incorporate spelling and grammar mistakes to remove the usual flawless presentation AI offers.

So what is this chatGPT AI all about? If Elon Musk was a co-founder of its OpenAI roots but appears fearful of it now and calling for regulation, it must be significant. Co-incidentally, looking at LinkedIn recently, I see that there are now rankings for the best AI software, such as:

ChatGPT           Solves anything
Writesonic       Writes anything
Midjourney      Generates art
Replit                  Generates code
Synthesia         Generates video
Soundraw         Generates music
Fliki                      Generates TikTok feeds
Starrytars         Generates avatars
Slides AI            Generates PPT’s
Remini                Edits pictures
Pictory               Edits videos
Wordtune         Summarises notes

I decided to take immediate action to see for myself. In my current volunteer role with Derbyshire Dales Community Energy where we are trying to use community funded rooftop solar to reduce energy costs by supplying low-cost electricity to local businesses, schools, public building and others, with no capital outlay, I am involved with some projects that need written overviews, reviews, feasibility studies etc as a pre-curser to funding applications. To summarise the reams of documents and reports into an overall overview was going to take some time. There are also some significant climate change challenges on the near horizon for our local World Heritage Site, the Derwent Valley Mills. My view on volunteering is that a volunteer’s time is a valuable asset and anything that makes their involvement more time efficient is of interest. Seemed perfect for chatGPT!

So I signed up for chatGPT and very quickly got to the point of asking it a question. I asked, ” What are the risks to the Derwent Valley World Heritage Mills due to climate change?” Within seconds I was looking at a page of coherent text outlining risks due to flooding, erosion, extreme weather events, temperature and humidity fluctuations, landscape changes and tourism. It offered advice on preventive measures and who to involve in their implementation.

Somewhat impressed I asked “What are the best practices for the protection of world heritage sites from the effects of climate change?” As before, within seconds I had a 10-point plan. The ‘conversation’ continued by my asking “Specifically what are the risks to Cromford Mills?” and then moved to funding, asking “What is the best source for funding to protect the Derwent Mills World Heritage Sites?” and in response to chatGPT’s reply I asked, “Of the philanthropic options, who would be the best?”. In reply chatGPT was suggesting that I research or assess various possible funders, so I turned its reply into a question and asked chat GPT to research and assess the potential sources of funding – and it did.

The full ‘chat’ can be found here.

From this, my first baby-steps of interaction with AI, it is clear that chatGPT AI is an extremely powerful entity that is able to understand the questions that I asked and instantly provide meaningful and accurate responses that align with the answers that I had independently arrived at through my own research and reading. It understands that its reply can raise further questions and as such you have a ‘conversation’ by simply typing your next thought or question without repeating the original subject in the question. It claims to ‘break down complex concepts into easily understandable language which can help in bridging a knowledge gap and that by democratising access to information it can empower individuals, regardless of their background or expertise, to engage actively in their topic/ project.’ It appeared that the only limit was my ability to think of another question.

I would conclude that;

1. chatGPT is able to understand the question exactly, and to add further questions into the same chain of questions and answers
2. chatGPT is able to accurately and quickly retrieve information pertinent and relative to the question asked and present it in a clear and easy to read and understand format
3. chatGPT ‘s speed of retrieval of accurate, relevant and suitable information is both impressive and overwhelming. It can reply almost faster than you can think of the next question.
4. chatGPT provides emotionless responses, they are facts and often state in a reply “As an AI language model I don’t have a personal opinion or access to real-time information”.
5. chat GPT reminds me of ‘Spock’ , the fictional character from the TV franchise ‘Star Trek’, known for his analytical mind and logical thinking
6. chatGPT’s speed of accurate retrieval of relevant information and reply can be imposing, overwhelming and capable of making a human feel cognitively inferior.

I believe that the challenge is how best to take positive advantage of AI technology for the benefit of mankind, use it’s logical prowess to offer solutions to some of the difficult questions that we currently face, and to see if it can assist in bringing balance to the world in both humanitarian and planetary terms. I feel certain that the publicly available version of chatGPT, is likely to be only the tip of the iceberg of what it is capable of, and that higher level of AI, be it chat GPT of by other developers, needs to be regulated and controlled.

I also used chatGPT to ask about the use of chatGPT in Community Energy: the full ‘chat’ can be found here.