The dark side of ChatGPT
It's been a while since ChatGPT and subsequent language models came along and "revolutionized" everything. I too am thoroughly impressed because it is an amazing brainstorming buddy and it can make your writing quite eloquent to say the least which sometimes is the goal but more importantly it can write code, social media posts according to platform (maybe not threads), essays, questionnaires and tutorials and who knows what more (a lot of people do). Which is cool right? Or is it?
Honestly the more I use it the more it horrifies me, not because it'll take away jobs but because of possible data contamination if that is the right word. Anyways let's jump right into it.
First let's talk about the good aspects of chatGPT:
It saves time and money on trivial things that you really don't want to spend time on such as boiler plate emails, slightly editing content between social media platforms, summarizing text, appropriating content according to personal or audience preference, throwing out random (apparently) ideas, rewriting poorly written content etc.
If you ask it what to use it for, it actually provides a decent enough answer, and it would decline answering certain sensitive questions on the first attempt which is a good starting sign.
It can provide boiler plate code in most languages and an intro to most topics in your desired format i.e. bullet points, paragraphs, visual explanation or analogies although the analogies are terrible (and so are the jokes).
All of the above make it quite a useful tool to have as long as you're not putting sensitive data on there and not using it is missing out.
Now that we've established that it's quite the batman of platforms, let's look at some really glaring flaws that can be quite dangerous.
As they say: "better go without medicine than call in an unskillful physician", which applies to ChatGPT or more importantly bad usage of it.
So let's start that list:
Call it irresponsible usage but AI produced text has made everyone an expert and platforms seem like the lawyer's equivalent to burying the audience in paperwork and the sad part is, it's not really verified, or even good content, it's just mostly word salad (i.e. the time when Joey from friends used that thesaurus).
It makes glaring mistakes and oh so confidently, where the mistakes range from simple coding errors, to bad references, incorrect facts. What makes this frightening is that since a lot of content sounds reasonable and at times it is factual, proper verification becomes less of a concern for some people in an era with small attention spans, FOMO and honestly too much to do.
As much as it helps creativity, it also hinders creativity by making the window of ideas narrower and narrower as we move forward. It's mostly spewing out the same content which may be true or false and a lot of people are picking it up and posting it which can create a bandwagon effect and a new data reality that just isn't remotely close to reality.
What to do???
So now that we'd established the good and the bad, let's see what we can do. There certainly are some actions that are paramount to ensuring this doesn't tarnish the already tarnished internet or communication as we know it, what with even more platforms.
Use ChatGPT for what it's intended use is, you can ask it and it'll tell you.
It would be a good idea for social media platforms to check for redundancies or accuracy in content posted, even though banning AI generated content maybe a bit more dramatic, not everything needs to be posted.
Be skeptical, when it comes to learning or answering questions i.e. trust more "official" content; courses and books even when ChatGPT seems right there.
Be aware of the bias within the algorithm, and avoid herd instinct a.k.a critical thinking.
Well that's mostly it and I'm sure I missed a lot of aspects but hey, you get the point.