Image generated by DALL-E for this conversation.
After winning a battle in 46 BCE, in what is now Türkiye, Julius Caesar spoke the well-known phrase: “veni, vidi, vici” – I came, I saw, I conquered. In November 2022 my son called me and said I should check out Chat GPT. That is how generative AI came to me. Following the hype curve, I saw what value and what dangers generative AI could have, but I started using it to test ideas, verify facts, create order from complex future scenario combinations and to assist me in structuring my thoughts. My first experiment with Chat GPT was to give it a heading outline of a 40 page report I just wrote for a client. This report took me four days to write, including the Internet research. Chat GPT wrote me a report that had approximately 60% of the content I used in 45 seconds! And then I realised, I will never conquer AI. To get the best of it, is to learn to live with it and communicate with it. Garbage in, garbage out is still true, therefore our conversations with our digital intelligence companions will have to be more precise, leading it into the logical thought we require to add value to our daily lives and work.
We should also be careful not to let generative be understood as generic. Here is a generic answer that Chat GPT-4 gave me when I asked the question: Is AI generative or degenerative for humankind?
“Artificial Intelligence (AI) can be both generative and degenerative for humankind, depending on how it is used.
Generative Aspects: AI has the potential to greatly benefit society in many ways. It can automate repetitive tasks, freeing up time for humans to engage in more complex problem-solving activities. AI can also assist in areas such as healthcare, where it can help in diagnosing diseases and predicting patient outcomes. In the realm of environmental sustainability, AI can optimise energy use and help model climate change scenarios.
Degenerative Aspects: On the other hand, AI can also have negative impacts. There are concerns about job displacement due to automation. AI systems can also be used unethically, such as in the creation of deepfakes or in surveillance systems that infringe on privacy. Additionally, AI systems can inadvertently perpetuate bias if they are trained on biased data.
In conclusion, whether AI is generative or degenerative for humankind largely depends on how it is developed, implemented, and regulated. It is crucial that ethical considerations guide the development and use of AI to ensure it benefits society as a whole.”
Generative AI models, like GPT-x (yes, I allow for further versions) or DALL-E (graphics from text), are trained on large amounts of data presented to it and learn to generate outputs that are similar to the training data. The key aspect of generative AI is that it is not just copying or reorganising the data it was trained on. Instead, it is using what it learned during training to create something new and original. This is why it is called “generative” – it is generating new content. While generative AI can create impressive results, it is still a tool that is guided by its training data and the instructions it is given. It does not yet have creativity in the human sense. It is not coming up with original ideas or making decisions on its own. It is simply generating outputs based on its programming and the data it was trained on. That programming though, is something that will develop and since AI can develop code, it may become more intuitive.
How should we adapt our way of speaking to generative AI? We have to give clear and precise instructions. It is imperative that we are critical in the reaction to those instructions. We cannot naturally assume that AI will be intelligent enough to place the instruction in exactly the same context we had in mind. We must be critical of the outcome. In giving instructions, we should note the following (says Chat GPT itself):
Specificity: Be as specific as possible. If you want a particular type of information or a specific task done, make sure to state that clearly.
Context: Provide enough context. If your request relates to a previous conversation or a specific topic, make sure to include that information.
Simplicity: Keep your instructions simple and direct. Complex sentences or jargon can confuse AI.
Patience: Remember that AI might not get it right the first time. You might need to rephrase your request or provide additional information. But in this context being right lies with the human being to make the final decision on accepting an AI suggestion.
Safety: Be mindful of the information you share. Avoid sharing sensitive personal information as AI systems store and process data. Do not provide AI with a full client report and ask for a summary. That report becomes part of the training database.
Ethics: Use AI responsibly. It is important to respect the same ethical guidelines in your interactions with AI as you would with humans.
The better you communicate with AI, the better it can assist you.
Can the largest degenerative aspect of AI be the erosion of the human ability to use its brain? Not thinking all the time, but relying on AI to solve complex problems for you may result in the de-evolution of your brain. It is like in athletics, you may be the fastest person on the track, but if you do not practice, you will not maintain that status.
There are two schools of thought on AI causing degeneration of the brain.
Cognitive decline
Over-reliance on AI could lead to a decline in certain cognitive abilities, for example, if we rely on AI for remembering information or for navigation, our own abilities in these areas might diminish over time. Our memory and spatial orientation skills might weaken if we outsource some functions to AI. AI might influence the structure and function of the human brain. For instance, the use of AI systems might lead to changes in neural networks within the brain, potentially influencing cognitive processes.
Cognitive enhancement
Generative AI can enhance human cognitive abilities, for instance, in handling large amounts of data and perform complex calculations quickly and accurately, which can augment human decision-making processes. It can also automate repetitive tasks, freeing up humans to engage in more creative and complex problem-solving activities.
The increased use of generative AI is unlikely to stop the evolution of the capabilities of the human brain. AI is designed to complement human intelligence, not replace it. It can handle repetitive tasks, process large amounts of data, and even generate creative content. However, it does not possess human qualities like consciousness, empathy, or the ability to understand context in the way humans do. The human brain is incredibly adaptable and capable of learning new things. This is known as neuroplasticity. Even as we use AI tools, we are learning how to interact with these systems, interpret their outputs, and apply their capabilities to new problems. The essence is to remain critical in what is suggested by humans and AI and to keep on validating. The evolution of the human brain will not be stopped abruptly, since it occurred over thousands to millions of years, a timescale that is much longer than the advent and rise of AI technologies.
In conclusion, while generative AI is a powerful tool that can change the way we work and think, it is unlikely to halt the evolution or development of human brain capabilities. Instead, it may shape these capabilities in new and unforeseen ways. It is an exciting field and one that will continually be evolving. We are together on this exciting journey to see how this relationship between human intelligence and artificial intelligence unfolds.
It is the same Julius Caesar that said: “Experience is the teacher of all things”. Both us as human beings and AI need experience to work together. So it is not about AI being generative or degenerative, it is about developing a language, thought processes and an understanding that will lead us into the future.
TechnoScene 