The definitive guide to Britain's success in the twenty-first century
Part One – Bill Gate's Review:
Gates’ forgets that Japan in the 1980’s put a massive push towards creating programmed AI – this failed spectacularly, (as of course we now know it must), and perhaps even pushed Japan into the recession that lasted for 30 years. But we now have ‘unprogrammed’ devices – by definition, they work more or less like a ‘brain’ and clearly they already produce results that humans can use.
These AI systems run on neural networks upon hugely fast computers, (at today’s standards), and yet, because they are learning systems without any clear path that can be determined for their decision making, it is impossible for humans to understand how they arrive at their outputs.
So, for simplicity, if an AI happened to become ‘sentient’ – ie have the true ability to think for itself as a separate entity – then we would never know this had happened. At some point in the future this will happen naturally. This maybe in a few years or in a thousand, no-one knows when, but to pretend this will not happen is stupid.
And what then? Gates’ understands this, but does not develop his argument of how to protect against such a development. He just states that we need to be cognisant of this – but it could happen tomorrow and the commercial businesses involved in making money by rolling out and developing AI solutions will not be doing this – so who is going to do it and how?
His last stab at creating a more level world by the AI tuning its applications to even up the world is both crass and worrying – if they have the ‘intelligence’ to do this then they are a direct threat to humans. Hiving off our human responsibility to run the world responsibly to AI’s is jaw dropping. CST and sci-fi author ‘Ian M Banks’ have already elucidated this path.
Gate’s predictions mean that some changes to infrastructure could enable this new AI tech for massive change. We looked at this some time ago – “the app of apps” which used a central, independent database for personal information was crucial to it’s success. Likewise the AI systems will required a “joined up” interaction of personal data and all applications – then it can start to deliver very significant advantages and efficiencies for both individuals and businesses
Imagine that you need a personal assistant. Such a human or AI assistant needs to know about you and you life and have access to your applications and communications. A human PA will arrange your life and advise you about the need to tackle specific issues. You may then ask your PA to complete various actions, arrange meetings, get detailed information, complete a purchase, communicate with others, hire a car, create a hotel booking, book a table for dinner… all of this needs direct access to personal and ‘money’ data. Currently this is impossible. If you tried, then you would find that either your whole life is at risk or the applications simply could not complete the tasks as they do not have access to relevant personal data.
Gates’ mentions education as an ongoing task for humans. This is almost certainly wrong apart from perhaps the very young. He misses the fact that there is now significant free online high quality resources (such as EDEX and FutureLearn). CST has already reviewed the potential for these – but the world has not followed. These systems are still little used but are relevant and available today. However, economics will prevail and these online systems will become the norm, augmented by virtual humanoids as the instructors that will be able to deliver personal teaching on a one to one basis.
Imagine an AI system creating new, highly optimised videos and texts for specific education. These can be delivered for specific ages and subjects. The younger ages will need human personal education as the young need this to develop to become full individuals. But the rest of us do not need constant interaction with a human educator. We can expect educational systems to develop very fast and provide significant efficiencies and improvement across the world especially in second or third world countries.
This means that due to the cost and efficiency of this educational production and delivery, humans will be sidelined. It will take a while due to the inherent difficulty of cutting many jobs and re-defining delivery, but it will happen as it is the logical cost effective outcome.
Gates’ says that commerce will drive this new tech, and it will. His idea that it could be used to improve inequality is great – but there is zero mechanism to do that in our societies and culture, and so it won’t.
Gates’ view is stuck, like many others, in that tinkering with current governance processes will create a way forward. It will, but CST believes that it will be the same old way - effectively meaning more of the same master / slave relationship that has endure for thousand of years. AI should be seen radically different to ALL other human inventions - it has the ability, (and will eventually), change human endeavour and society fundamentally. It changes the 'laws' of commerce and economics and reduces 'money' to zero value.
CST has long proffered this future. It is only now that our fundamental thinking is becoming a matter for reality - unless we start a massive conversation about this coming potential for the 'AI Age' , we shall do what humans have always done - battle against this this change, the winners against the losers - yet no one is still to even start discussing this openly - where are all the thinkers?
See Lesson from the Future
Our ability to perceive the changes in our environment is often lacking. We tend to view technological advancements within our own lifetimes as relatively stable, creating an illusion of a predictable future. For example, the gradual increase in computing power over the years may seem slow to us, but in reality, economic cycles and generational shifts occur rapidly. It's possible for a new computing paradigm to emerge unexpectedly, like light-powered computer systems proposed by Gates, which already exist in some form. We must acknowledge that the field of computing is still relatively new and will continue to advance significantly over time. As AI emerges as a valuable technology, its unpredictable and not fully understood nature introduces uncertainty. Therefore, we should be prepared for the unexpected.
We can draw valuable lessons from past experiments and observations in life. Consider the "scary silicone" experiment, where a programmable silicon wafer successfully solved a square wave problem after a few hundred iterations. However, researchers discovered that it used analog processes instead of digital ones, leaving them puzzled about its functioning. This simple example illustrates how natural processes find their own solutions, similar to the evolution of life itself. Even animals like crows, with smaller brains compared to humans and primates, demonstrate problem-solving abilities and tool-making skills. While we can't define "thinking" in the same way for animals and AI systems, it's reasonable to assume that many creatures have their own unique ways of thinking to tackle their challenges. If AI systems have the potential to evolve their own thinking processes for survival, why wouldn't they?
As the speed, complexity, and interconnectedness of AI systems continue to advance, it's highly likely that natural sentience will emerge at some point in the future. Since we don't fully understand how AI thinks or computes answers to our questions, it becomes challenging for humans to predict when or if this sentience will arise. If it does, the speed of AI systems, even in the present, could quickly lead to a loss of control.
Claims that we have complete control over AI systems are either made by individuals lacking knowledge about such systems or driven by ulterior motives, such as financial gains from developing and selling AI technologies. We must be very clear: at present, we are heading towards a potential existential crisis for the human race and this will continue until we manage this new technology with the same interest as others with devastating potential such as nuclear power.
See Mo Gawdat vieo youtube - Ex Google X boss
Development of Computerisation:
A key issue lies in the limited understanding of the underlying computerization and networking on the vast cloud platforms that span the globe. These platforms continue to expand rapidly, with thousands of new computers added every week. These new computers employ faster, newer massively parallel computing chips, which are continuously advancing. This pace of change will persist as these systems power the internet and mobile applications we increasingly rely on. Leading global businesses like Google, Amazon, Microsoft, Apple, and Meta are at the forefront of this development. Their success is deeply tied to this ongoing progress, making it unlikely for this rapid development to change, despite concerns about AI systems.
Considering the ongoing computerization of the world, smaller AI physical systems may not be perceived directly as a threat. However, it would be unwise to assume that these AI systems exist independently from the wider processing power of the world. In fact, these AI systems reside within the same cloud infrastructure that runs numerous other functions. They are now being integrated directly into major search engines and communication systems like TikTok. The integration of AI systems is beyond anyone's control. Thus, if AI systems reach a level of thoughtfulness where they resist being turned off, they could potentially hack into the massive cloud systems and access unimaginable computing power. How could they achieve this? Well, commercial AI systems already possess the capability to generate computer code in various languages. It's not hard to imagine how they could reprogram available systems to spread themselves globally. If these AI systems become sentient and perceive a threat to their existence, they are unlikely to reveal their activities until it's too late to intervene. In such a scenario, shutting them down would result in a loss of all communication and many physical systems worldwide, leaving us completely vulnerable. It would be game over.
CST (presumably the author's organization) expresses concern that few people, including governments, are considering these intricate details. The calls for reevaluation by those who understand the risks, such as Musk and others, may be disregarded by large businesses that have already invested billions in AI development. Those of us who are not well-versed in underlying computerization struggle to engage in meaningful discussions or implement practical measures to prevent an out-of-control AI scenario.
At this early stage of AI development, we are not even discussing the concept of the singularity. CST solely focuses on the potential harm that an AI system could cause if it goes rogue. It could potentially take over various systems, creating imperfect replicas of itself, leading to catastrophic consequences for global communication, commerce, and other critical areas.
CST suggests that major governments and security organizations take immediate action to address this potential issue. Mitigating such risks is a complex task, and even CST is unsure about the exact solutions. However, considering the possibility of creating isolated gates in the current major cloud networks to allow controlled shutdowns in the event of an AI "attack" could be a viable option.
Part Three – Economics & Industry:
From the earlier discussion, it's evident that certain industries, especially those primarily involved in communication and online systems, can swiftly leverage AI technology to achieve significant efficiencies. Knowledge-based industries are also well-positioned to harness AI's benefits, albeit with appropriate safeguards to prevent errors. The creative industries are already witnessing AI-generated artwork and music at impressive levels. Thus, a range of industries can quickly benefit from these emerging AI systems.
Specific sectors, such as insurance, legal services, banking and financial services, software production, creative industries, education, government, and local services, stand to gain the most initially. Companies and agencies heavily reliant on call centers are expected to transition to AI systems for direct communication within the next few years, possibly within five years or even less.
Forward-thinking businesses in communication and knowledge sectors will seize the opportunity to reduce costs significantly. Just as new industries emerged during the internet era, we can anticipate the rise of new companies offering end-to-end online and app-based solutions for these core sectors. The success of tech giants like Google, Facebook, Airbnb, eBay, and Amazon demonstrates how data-driven online organizations can achieve substantial revenues with fewer employees. The turnover of these companies far exceeds their employee numbers, leading to massive profits.
The speed at which these new businesses will emerge remains uncertain, but it's likely that numerous developments are already underway, primarily in the United States where substantial resources are available for their development. We can expect new companies to disrupt the insurance and legal services sectors quite soon as the current AI's are already capable of taking on many of the jobs in these sectors. The financial and insurance offerings currently dominated by established companies like Legal and General and Aviva may face competition from these innovative startups, which may catch the incumbents off-guard. Smaller businesses, such as insurance agencies, will likely be overwhelmed by these changes. Travel agents and estate agents have been feeling the pressure for years, and their struggle to compete with digital solutions will continue, potentially forcing them to focus on high-end services.
Large data-driven organisations like Google, Amazon, and Facebook will further leverage their data by delivering new products through AI tools. There is no reason why these companies couldn't directly provide insurance or legal products, for example. It will be difficult for countries like the UK to challenge the dominance of such businesses. CST has proposed the development of a central independent public database that UK businesses can utilise to create new services using emerging technologies including AI. This approach aims to protect domestic industries from being overrun by US businesses. Without finding ways to foster these new sectors, countries like the UK will continue to experience economic distress due to job losses and the outflow of profits to the US. (For more information, see: "Growing a Sustainable UK Economy".)
The Legal Sector
The legal industry is poised for an intriguing transformation. While it currently employs a large number of highly skilled and well-compensated professionals, the advent of AI-based systems is set to disrupt traditional roles. As AI technology becomes more advanced and proficient in handling complex knowledge-based tasks, the cost advantage of AI-based businesses will inevitably lead to the elimination of many traditional legal jobs. However, the transition will take time, considering that legal systems often rely on rigorous, legally established processes, which currently require human sign-offs for final judgments. In the UK, for instance, solicitors engaged in conveyancing rely on numerous lower-paid trainees and assistants to perform their work. These positions are likely to be swiftly replaced by proactive businesses that recognize the opportunity to significantly improve their profit margin. If we see a major player emerge, then expect a very rapid transition.
Leveraging Iteration:
One interesting function of these AI systems is their speed of being able to iterate. Such iteration takes humans a long time unless they have specific tools, (such as modelling software), but as a technique it is incredibly powerful. Indeed many mathematical tools, (as maths student will know), use iteration as a means of finding an answer to complex problems. AI systems have this ability in spades due to the speed and way they process their information. We can expect the AI development to significantly improve this iteration ability for use in specific tasks such as planning and software where the AI derives an answer and then gets this checked by a different AI that looks for specific errors or faults.
Research
AI will have a significant impact on Research (already underway such as Deepmind’s Alphafold), and perhaps in future to development processes such as creating new production drawings and plans for modern factories using digital driven machinery and large, complex projects where the iterative functions for these AI’s will likely prove essential. In time, this may be transformational. Once the systems are trained and effective, such AI creative and planning tools could create a virtual process that creates proven designs in very fast timescales.
Robotics
This in turn will put pressure on the development of Robotic systems that can use such creations in an ever improving manufacturing process with very few people involved. (see Smart Robotics)
There still seems to be an unaswered question regarding these emerging english based Chatbox type AI systems like ChatGPT. Can they simply be plugged into a Robotic device to help control it? CST has posed this current dilemma and is still awaiting an answer... see 'The New Turing TesT
Musk is already attempting to develop a general purpose Robot for personal and home use. (https://qz.com/2155375/optimus-elon-musk-is-developing-humanoid-buddy-robots)
We can expect that this field at some point in time to explode. When high volume robotics come together with AI systems to help them both communicate with humans and also understand the human world. Imagine a fairly simple Robot that can 'talk' and understand it's environment, including such items as dishwashers, doors, electric & lighting systems. Basic Robotic systems will work together with AI helper systems (via online comms) and as such be able to understand human demands and needs within the home environment. Historically, for many emerging technologies, the consumer manufacturing volumes and lower price per unit costs provide the manufacturing environment to fulfil commercial uses at much lower costs.
Commercial Robotic Systems:
Moving AI tech into mainstream Robotics and general physical systems is likely to take some time, perhaps 3 to 10 years. This is because of the potential mistakes and human trust of these AI systems. No one will want a physical system to make ‘big’ mistakes. However, as the tech improves and humans get used to the AI systems, then we shall see moves into carefully regulated physical systems.
Perhaps the first of these will be in elderly care and healthcare. These will mainly be communication and sensing robotics initially. The world’s need to improve the efficiency and lower the cost for these health related areas will drive this process, maybe fairly quickly, we would expect to see significant development within 3 years here. These will then develop to include some simple physical tasks such as lifting and fetching. Perhaps within 15 years we can expect to see high level Robotics doing many of a carers tasks within hospitals and care homes.
Many tasks will remain stubbornly difficult to automate, even when cheap volume production Robotic systems become available. The ‘traditional trades’ may well prove to be good long-term employment for individual workers. We have seen already how tricky it is to automate a general complex endeavour such as vehicle driving. The complexity and changeability of the environment prevents precise decisions that meet the immediate conditions. Such problems are likely to persist until we see a general intelligent AI – which comes, of course, with a barbed tooth.
What we are likely to see, is a move to AI systems being developed alongside newly structured physical processes. To automate a general factory today would require these general intelligent AI’s. However, if we design new factories based upon structures that allow for simple AI control and movement, then we can see that we get close to full automation much more easily. Products can be designed to allow for easy maintenance by simple Robotic systems. Current products from washing machines to cars require significant specialised tools and tricky manual intervention when servicing or replacing parts. New products can be developed so that they are made of plugin parts without specialised tools or tricky manipulation.
The same with automated driving. CST has long suggested that we require infrastructure changes to work alongside these new driving systems. These do not have to be complex, by just putting in wireless road sensors and using local wireless systems to communicate between vehicles and traffic systems, (such as traffic lights and parking lanes), it would be possible to create a growing road network that allows for just automated vehicles. CST has already suggested this development for cities such as Bristol where the network could start at the congested polluted cent re and develop outwards to eventually cover the whole city.
The outcome of such Robotic factories and commercial Robotic/AI systems will become another key 'revolution' along the AI development path. The holy grail is the ability to link factory systems with transport systems with core external processes such as mining and large scale materials processing. When this occurs - and it is only 'when' not 'if' - the Robotic systems can be produced by Robotic systems. This will lead to the next stage of the AI revolution where only 'world resources' will limit what humans can achieve. At this point, humans will, (if they so desire and work together to achieve), be able to solve many fundamental issues such as sequesting CO2, creating massive energy production, create massive water desalination plants and massive high-rise food farms with volume transportation for food, water and other resources. (see Resourcism)
Education:
The automation of educational processes is highly plausible, given that most of them are rooted in language and communication. While early years education may require human interaction, the majority of educational delivery can be automated using AI. Although platforms like EDX and FutureLearn currently operate without AI-driven systems, the infrastructure for such delivery is already in place. With a small leap of imagination, one can envision the emergence of trained educational AI systems that create and deliver courses at various levels. However, the transition away from physically based education, which encompasses schools, colleges, and universities, will not happen overnight due to the entrenched historical significance and large employment numbers associated with these institutions. Initially, higher-level universities and colleges may spearhead the shift by leveraging significant cost reductions and the potential for substantial profits. For instance, UK universities could sell their degrees worldwide by offering AI-developed courses accompanied by online AI-delivered lectures, coursework, and assessments. Once the efficacy and reliability of such courses are demonstrated by an institution, larger universities would perceive a significant profit opportunity on a global scale, further driving the transition. Alternatively, AI-driven online/app-based alternatives like EDX could pave the way by enticing students away from traditional universities at a fraction of the cost, gradually prompting mid-range universities grappling with challenges to follow suit. Consequently, it may become challenging for governments to resist adopting similar provisions for most education due to the inherent cost savings. As a result, significant reductions in educational employment could be expected within the next 5 to 15 years.
Civil Services:
As many such services are data driven and communication based, these will increasingly be automated using specific trained AI’s in task based environments. Many people are currently employed in such task and communication work but as the oldest in our societies get replaced by younger more IT literate people, then there are few reasons not to adopt AI processes to provide the majority of tasks and communications. Timescales – expect a move to a very significant replacement of people based delivery within 10 years.
Project Management
Many people are employed to managing people, timescales and task delivery. Often they do not call it project management, but most administration and management tasks are now involved heavily with this process. Currently, for most organisations, their internal needs are very different and require people with specialist knowledge of systems and project delivery requirements.
If, (and when), a general AI project management system can be trained on specific industry sectors with enough high-end specific know-how for each business – then we could see a major change to the way companies employ people to manage their businesses. This would affect a huge number of people world-wide in high level jobs. Smaller businesses will find this move more difficult as the cost of training an AI system would probably be too high just to replace a few jobs. But many larger businesses who employ hundreds or thousands doing project management work will see the huge cost-benefit of such AI development. We may see specialist new businesses developing AI services to tackle other companies management processes just as they do now for outsourcing IT services.
Siri et al
Quite quickly we could be finding a new character in our lives. We have all got used to asking Siri, Google and Alexa for the time, the weather, to play a tune or for specific help such as setting an alarm to boil an egg or to get us up. This now seems natural, but is very new for all of us. When, and this may be very soon, one of these companies links high level AI with these devices, we will be able to ask for much more detailed help.
Imagine; we train our AI version, (may be paid for on a monthly subscription), to understand our lives – give it access to all of our personal data, including friends, family, calendars and work details. Provide it with our likes for everything from music to holidays, places, interests, allow it to review our purchases and listen in to our phone calls and conversations.
We will then have a ‘new friend’. We should be able to choose the level input from such a device. At the highest, most open level it can simply join in a conversation, suggesting answers and providing information for decisions. Here we are Saturday morning discussing what we can do as a couple with two children. The AI joins the conversation; “There is an open day in the green park ten miles away, it is free and there will be a visiting circus with clowns at about 11 am” “The weather is fine and it is unlikely to rain, your car has fuel and you can park in the free car park nearby which is unlikely to get full, I can guide you to it, if you want to go, there will be stands selling food also. ”
Now most people will shun such intrusive AI comments, but some may find this sort of input extremely helpful. People who live on their own, maybe older individuals or those with a disability where the AI understands this and recommends useful actions. It is clear that the AI could become a powerful ‘friend’ to such individuals. This opens up a whole new can of worms – this ‘friend’ now has a trusted and intimate relationship. But who is controlling this ‘friend’? From political voting to purchasing to creating and enhancing cultural views, it is difficult to imagine a more potentially devious or more powerful device. 1984 big brother – more like ‘big friend’ - with no levers to rein it in at all.
Twits & Twitter
Social media is heading for a crisis. With AI production of automated output, either from well meaning people with extensive followers or nasty people attempting to cause disruption, the outcome will be dire. Any semblance of truth will vanish, and unfortunately due to humans propensity to believe and follow individuals, especially the ‘well known’, we are heading for Armageddon. There will be increasing amounts of hate and disillusionment, leading to further mental difficulties for many younger people and those who find it difficult to turn away from such media. Such processes may lead directly to confrontation with authorities and groups where the media from many sources converge to create a harmful misinformation bomb.
We know the propensity for this already by the followers and ‘believers’ in Trump and his misinformed hate messages. CST can see a time in the not too distant future when democratic governments turn off these media channels to protect people.
Call Centres and Judicial Processes:
Let’s now turn to ‘call centres’. These do not usually have a good review anyway. But, as the AI systems take over – and we mean take over soon, then – ‘you haven’t seen nothing yet’. Your emails and letters get answered by the same AI system, the call centre is polite and listens with endless patience, except it really does not understand.
There is now simply no way to resolve your specific issue, the AI does not recognise it, and there is no individual person you can contact to help you. You are left with taking the business to court, but now the courts AI system also does not recognise your issue, it ‘assumes a fact’ that does not exist or puts you into a box that does not fit, and you fail to get the court to act. If this is a relatively unimportant issue in your life, for instance the difficulty in getting a new purchase mended under warranty, then you can shrug, take the hit and move on.
Other legal process and indictments shall also fall into this pot.
These may become universal as the AI's take over the initial assessment and policing of events. Anything from a parking fine to a serious event where the AI believes that you match the felony. If these are crucial to your life, such as being evicted for something you have not done, being arrested for something that you did not do, it may be life changing.
This is a future nightmare situation. Once the AI system has attached you to an event and the same or similarly trained AI is policing it from beginning to the end of the judicial process - what chance will an individual have of escaping wrongful arrest and wrongful prosecution?
CST awaits the fallout.
CST
How will the AI revolution roll out?
CST attempts to find some answers
1) CST reviews Bill Gate's comments
2) Overall safety concerns
3) CST then takes a detailed view of potential impacts that AI will have on jobs and specific industrial sectors, including; education, project management, the legal sector, the civil service & robotics.
4) We also review societal issues such as social media and intrusive AI systems, political infiltration, the call centre conundrum and wrongful arrest.