AI and Big Data

Naturally Smart Writing, Beyond the Artificial

A human framework for content writers

This piece is about proposing a framework for smart human writing. There’s such a flurry of social media content on AI one can legitimately feel bewildered. Every day we are overwhelmed with new GenAI tools, all deemed more “revolutionary” than those we had the day before. But do these tools really make us more intelligent, or on the contrary are they a way of outsourcing the intellectual work we did ourselves in the past? We asked ourselves here what would be the characteristics of intelligent human writing, as we promote it within our 100% human-generated content initiative.

Smart Writing Rather than Artificial, a Framework

Smart Writing by humans is possible
Smart Writing by humans, such is our proposal with this framework

Not So Smart: AI Confusion, AI Overflow and AI Overkill

In the past few days, LinkedIn AI gurus have become more insistant. ChatGPT 4 is revolutionary, but Copilot is already sitting in your mobile, nope, forget about it! Go to “Le Chat” (I suppose one would need to pronounce this as in “Shah”), unless the future is with SGE, Gemini and Google! Oh no! Google is going to crack down on AI content (I’m slightly dubious on that one), not forgetting the “Copilot is a must-have” (another one, not the same one, the one from Microsoft 365, utterly confusing if you ask me) and last but not the least “Hell no, the future is Perplexity.ai…”

These and many other tools are well known to us. Most of them are on our computers, tablets or phones, and we use them every day. We know them and most of their nooks and crannies. We have trained hundreds of young people on these subjects (and asked them to look at all that with hindsight). However, how much of too much is overkill? Doesn’t too much AI kill AI? Incidentally, I notice that when you dictate ‘AI’ to macOS, it still gets it wrong 9 times out of 10. Maybe someone could start by fixing that…

Ultimately, as we are piling up dozens of tools are we getting “smarter?”

We raised the question of what is intelligence a few years ago, at the start of the new wave of AI popularity in 2017-2018. More recently, we were pleasantly surprised to see it taken up in a more sophisticated and elaborate, and above all much more erudite, form by French AI expert, mathematician and philosopher Daniel Andler in his opus artificial intelligence, human intelligence, the double enigma, published by NRF, 2023 [not yet available in English, sorry].

The question of the definition of intelligence is endless. There is no point in trying to resolve it in this piece. But still… Can we assume that human intelligence is limited to problem solving? Certainly not.

For this mini-post, I asked myself the question of metaphorically representing the characteristics that seem to me to be the most important when it comes to Web content strategies, if I want to stick to my field of practice. So here’s what I’ve come up with. This, for me, takes the form of an anagram, C.L.E.V.E.R.

Here it is with a picture that’s worth a thousand words, even though ChatGPT is so wordy that I’m sure it would disagree with that statement.

Smart Writing by humans
This drawing is guaranteed 100% human with 0% artificial intelligence flavouring and no conservatives. It was made by yours truly with his bare hands and a stylus using the Paper app on an iPad. Smart drawing for smart writing. Makes sense.

A Smart Writing Anagram

  1. C for Cunning: LLMs might be quick but we can outsmart them. For example, one of our friends has a newsletter and she has a trick I really like. When she needs to cover a topic in her newsletter, she goes to ChatGPT to make a draft… And she decides to contradict it, using ChatGPT as if it were a sparring partner. This is a vulpine approach to authenticity (from vulpinus, Latin for fox an animal renowned for being shrewd) without rejecting AI but being smarter. Simple but strikingly powerful!
  2. Lateral thinking: a term coined by Edward de Bono, to represent the ability to link seemingly unrelated elements together and produce new ideas. A process some call ideation. Indeed, it’s one of the traits of human intelligence that can sometimes be linked to solving a particular problem, but not always. This is where creativity lies. Some will retort that LLMs are very good at suggesting out-of-the-box ideas. My experience showed that these tools can sometimes serve as a starting point when you have a bout of writer’s block. Yet, solving the equation yourself is more rewarding. And above all, it encourages us to think. Let’s take an example. I can solve my weekly cryptic crossword puzzle by looking up the answer in the online crossword dictionary. But it’s a lot more fun to rack one’s brains for two days trying to get to the end with pen and paper (or stylus and tablet in my case). In fact, thinking is what keeps our brains working.If we stop thinking, we’ll end up living in an idiocracy.
  3. E for experience: experience is something GPT or its many clones will always lack. Because they will never be going out into the field. What’s that for human superiority? A good pair of legs.
  4. V for visionary, that’s very apt for us. Humans are keen to project themselves, sometimes very badly, into the future. In fact, it’s one of our favourite sports. Just watch the BBC and you’ll have your fair share of experts in anything talking about anything with aplomb. And yet we listen to such experts with great passion, because there’s something fascinating about trying to understand the future. Even though we can’t. We could, of course, ask ChatGPT to perform its usual stochastic tricks and mimic our experts’ gibberish. But LLMs are only looking back. We, can project ourselves into the future. This is most probably a futile exercise, but humans love it anyway and always will.
  5. E for enlightened. The question of the use of LLMs is also a matter of philosophy in relation to our link with culture and information. Curiosity, the incessant search for information on anything and everything, is certainly one of man’s defining characteristics. Admittedly, this is a point that is suffering these days with the introduction of response engines in lieu of search engines. Nicholas Carr pointed out a long time ago that  Is Google Is Making Us Stupid and now things might get completely out of hand. All the more reason to actively combat this urge to get an answer before searching. One needs to document and research our topics and those who do will be the winners.
  6. R for rigorous: this is certainly one of the most important points. The more language models hallucinate, and we’ve seen some recent examples related to Chatbots, the more we need to get our critical minds working, cross-reference information, check it out. We must never give up.

So here’s our little funny acronym. We have no great ambitions for it but are merely trying to set the record straight. Let’s assume it could serve as a guide to defining your human online content strategy.

I can already hear naysayers pointing out, as has already happened, that not all humans are intelligent, and they would be perfectly right. Not to mention that many, in the past, the present and, alas, the future, have used and will use their intelligence for nefarious or downright wicked purposes. It’s true that human intelligence isn’t always perfect. We also wish that algorithms aren’t trained with this background in mind. Let’s focus only on the bright side of things, it’s a bit unsophisticated, I agree, but it’s a lot less stressful.

 

Yann Gourvennec

Yann Gourvennec created visionarymarketing.com in 1996. He is a speaker and author of 6 books. In 2014 he went from intrapreneur to entrepreneur, when he created his digital marketing agency. Yann Gourvennec a créé visionarymarketing.com en 1996. Il est conférencier et auteur de 6 livres. En 2014, il est passé d'intrapreneur à entrepreneur en créant son agence de marketing numérique. More »

Leave a Reply

Back to top button