I have been reading a number of articles about AI and many talk about the risks. One in particular which I had mused over before... just not in the context of AI. I thought I'd share my musings but please note, I'm not an AI expert, I'm just sharing some interesting thoughts of people who are, along with my opinion.

First, lets start by defining some terminology

Artificial General Intelligence

Artificial General Intelligence (AGI) refers to highly autonomous systems that possess the ability to outperform humans at most economically valuable work. AGI systems are designed to understand, learn, and apply knowledge across various domains, similar to the cognitive capabilities of human beings. Unlike specialised or narrow AI systems (the only type of AI we have today), which are designed for specific tasks, AGI aims to exhibit intelligence and versatility comparable to human intelligence across a wide range of activities.

AGI is characterised by its capacity to comprehend and learn from diverse information sources, adapt to new situations, and apply knowledge to solve complex problems in a manner similar to human cognition. It encompasses not only specific tasks but also the ability to transfer knowledge and skills from one domain to another, exhibiting a form of generalised intelligence.

Achieving AGI involves addressing a multitude of technical challenges, including perception, reasoning, and decision-making, among others. The realisation of AGI could have profound societal implications, revolutionising industries, economies, and the overall human experience. Critically we have no idea how to create AGI

Source: Forbes

Generative AI

Generative AI is a type of artificial intelligence that can create new content, such as images, text, or music. It learns from huge amounts of existing examples and patterns in data to create models of how to generate content (known as large language models or LLM for text-based content). It then uses these models to generate new content that is similar in style or feel. In other words, it simulates human creativity and produces novel outputs. Generative AI can be used in various applications, from generating realistic images to composing music or creating unique text. It allows machines to be creative and generate content that was previously thought to be exclusive to humans. OpenAI's ChatGPT and Google's Bard are both Generative AI.

The less understood risks of generative AI

Is Generative AI (or ChatCPT) sentient (the ability to have conscious experiences and exhibit subjective awareness)? Well, everything we know about sentience is incompatible with a large language model, which lacks any coupling with the real world beyond our text inputs. To experience sentience, one needs the capability to perceive and engage with the complex and ever-changing world using sensory organs and a connected nervous system that is intertwined with the physical environment. Without this connection to reality, there is no sensation, no comprehension, no reality, and no stable sense of self. So no, ChatGPT is not sentient or AGI but it does still propose some risks (not that sentient AI is specifically a risk in itself anyway)

We have all read some headline about about the existential risk of AI or even how AI is causing real problems right now. From algorithmic discrimination and disinformation to growing economic inequality and environmental costs. Whatever amount you think these are affecting us today or not, they are clear and real risks that most AI experts accept, but they are also somewhat understood with lots of discussions happening about how to apply guard rails for them. Something which is not, is how AI's positive qualities may also affect us negatively in the future.

We are now in an age where machines produce output that’s often ‘good enough’ to pass for the labours of our thinking. What's concerning is the increasingly common argument that thinking is work we should be happy to be rid of. Evolutionarily speaking, thought is expensive, so our brain is tuned to be lazy and is conditioned naturally to not want do any work unless it has to. So it's easy to see how this could become a bigger issue. In fact, with many major technology shifts over the last few decades, we have seen society shifts like this. Most recently we've seen this out of social media where people go looking for quick answers and believing any random statement coming from any random Instagram video from @stevecorgel, without thinking it through or verifying it because @stevecorgel's world view simply matches their own. Then re-sharing it until it becomes a more widely held belief, even though it's completely wrong.

Source: AnalyticsVidhya

There is hope. Some experts are researching how the underlying negative consequences of positive values of generative AI risk creating societies where intelligence, even thinking in general, aren't valued. Risks of a future where AI is creating a dull, un-innovative and uninspiring life, where human existence regresses, with little creativity or dreaming because AI is simply repeating its findings without sensation, comprehension, reasoning and ethical decision-making. Perhaps even risking opening of doors to easier manipulation of popular thought by people in power. [Enter Bond villain, stage left]

We use AI... I use AI... Some of this article was created with the assistance of AI! I absolutely see the huge benefits of generative AI in workload reduction, but I come from a generation which didn't even have the World Wide Web and therefore have a precondition to still think through anything it creates. I'm not so sure younger generations will do the same. I don't want that and I certainly don't want that Black Mirror style future. I vote for ethical and responsible AI.

#ethicalAI #responsibleAI