Accounting for Humanity

This is a brief thought (maybe, I say that a lot an then write an essay) that quite regularly passes in and out of my mind. When we talk about Artificial Intelligence, how intelligent are we actually talking?

Now, as a disclaimer, I am not referring to the highly advanced, technical wizardry type AI’s that do in-depth data/statistical analysis. I am looking at this more “public” AI that is seeping into our everyday lives (I swear EVERY software I use now has an AI assistant that has somehow managed to make itself more annoying than Clippy). I am also not suggesting that I am an expert on the inner workings of AI (I am undoubtedly, and ironically, misinformed to some degree about how it works), but my basic understanding of how we as as society are using or interpreting it is a valid opinion.

Stupid is as Stupid does

My (currently) passive concern is the reliance and belief society is building that AI is infallible, perfect, superior. We seem to overlook that GenAI or these public open access variants do most of their learning by trawling the internet. The same internet laden with misinformation, conspiracy theories and harmful discourses/narratives.

Thus, two issues arise here. First, if we blindly believe what AI tells us, there is a good chance we will be living our lives hugely misinformed. Second, if we are thus living with half truths and fabrications, it is likely that any content we “create” (i.e., if we even create it at all or use another AI to do it for us), we further dilute the quality of information available to the AI to learn. So begins a downward cycle of Artificial Intelligence becoming Artificial Ignorance.

So…Are we the problem?

Before going any further here, lets recall Tay, the AI-powered Chatbot that went from a novel idea to a racist, antisemitic nightmare (https://www.bbc.co.uk/news/technology-35902104). This highlights the problem that (attempting to give benefit of the doubt) there is a lacking awareness that plying AI’s with harmful or frivolous (deliberately false or comic information to confuse it) can lead to valuable and beneficial information being degraded if the volume of negative information is greater.

Think for example how children learn. At the moment AI is learning from us thus, we need to consider that we hold a degree of responsibility and accountability toward what we choose to teach it. Furthermore, there is also perhaps a need to reconsider how we view AI. It is the 21st century rendition of the printing press i.e., this a technology/tool to give us access to new sources of information to enhance our learning and knowledge NOT to supersede or replace it.

It’s a marriage not a war

So what is my point here? For lack of a better analogy, AI is not something we should fight against or seek to destroy but it also is not our knight in shining armour. We both need each other to do our best if we are going to succeed. As a human you possess the gift of critical thought, hold on to that, you are going to need it. Stop thinking that AI knows everything and (more importantly) that it is right about everything.

We are at a critical stage in our (and AI’s) life, we can work to make both of us reach the next level of knowledge or, our lack of self-worth can lead us done a dangerous path where we think all though extra limbs in the photo’s are real…