Why write another article about ChatGPT? When you cut through all the shouting and sycophantic praise, the advent of tools like ChatGPT and Dall-e-2 are forcing us to question what it means to be human and open the door to more radical views. It’s as if ChatGPT is lifting a veil and forcing us to question many foundational things about our modern lives.
There has been so much written about ChatGPT, that I felt that adding my voice to this discussion wouldn’t really make sense, but over the last few weeks I have started to change my mind and this article is a product of my current thinking on the topic.
As everyone else, I warmly embraced ChatGPT when it came out and enjoyed playing with it. Getting to understand how to prompt the algorithm to get the right output was quite thrilling particularly when the output was so eloquent and credible. I used it for a few things like writing a formal letter to my daughter’s school (which I ended up rewriting almost entirely) and helping my wife write some business emails.
So, what is this technology that has created this level of interest? Let’s get ChatGPT to answer itself:

Figure1: ChatGPT in introspective mood
This is a new kind of chatbot using the latest AI technology allowing it to generate human-like responses to prompts from users. The chatbot has been trained on a huge amount of data taken from the internet up to the end of 2021. The size of the data used to train this algorithm is quite mind-boggling and includes the entire internet content collected over 8 years of crawling data, large collections of books, all Wikipedia articles in English and a collection of specific web pages considered to be more authoritative[1] .

Figure 2: The training process. Source – OpenAI
There are many sources explaining the way the algorithm works so I will not spend too much time on it. The simplest way I can describe it is to say that through clever maths and the large set of data it has been trained on, the algorithm can predict a sequence of words based on a prompt provided by the user. The beauty of ChatGPT is that it can do this in a conversational way keeping a memory of the prompts and producing more accurate results as it uses the feedback to improve the word prediction. What makes this mind-blowing is that the results are of extremely high-quality and can adapt to any context. ChatGPT can write about ethics, translate languages, write software code, and create scripts for plays, amongst the many experiments people have been using it for. The quality of the generated text and the flexibility of use cases is what has caught people’s attention and triggered the inevitable wave of comparisons between the output of ChatGPT and what humans can produce, including some very interesting experiments.
This is where I want to take a detour from talking about ChatGPT and its incredible potential to tell you a story about a night in Santiago, Chile in 1996…
I remember meeting my friends early and getting to the venue hours before the start of the concert. Chile had come out of dictatorship a few years before and more artists were now including a stop in Santiago (Chile’s capital) as part of their world tours, so we got to see artists like David Bowie, Peter Gabriel, Metallica, Guns and Roses[2] and a lot of other legendary musicians of the time. Together with massive stadium rock events like these, smaller bands were also touring, and I had been excited for months because Nick Cave and The Bad Seeds were part of the line-up of a small festival called Crazy Rock with a few local bands and also Love and Rockets and Cypress Hill, of ‘Insane in the brain’ fame. Nick Cave had just released the album Murder Ballads and I couldn’t believe I was going to see him perform those songs live in my own country. There was a strange mix of pride and anticipation amongst us as we got the best seats in the house and waited for the concert to start. Trying to convey the excitement and madness in that place will take a very long time and this detour is already losing me a few readers who wanted to get my take on ChatGPT. I promise I will come back to it.
The audience was at breaking point by the time Nick Cave came onto the stage, he started walking around, talking to the audience, and jumped into the first song. Instead of the release and enjoyment I had anticipated, it all quickly turned into a nightmare. The crowd at that point just wanted to see Cypress Hill and had no time for Nick Cave’s sophisticated renegade rock. They started spitting at him and the whole atmosphere became very charged. Following several warnings and songs interrupted half-way, Nick Cave and The Bad Seeds played only a fraction of the songs they had planned to play and abandoned the show. We were all speechless. The sense of frustration and shame was intense, and I remember feeling angry for days after that.
Fast forward to 2002 and I was already living in London when I saw Nick Cave at the Royal Festival Hall and I could finally close a very long and bitter chapter in my musical history. I have continued to listen to Nick Cave through the years and consider him one of the greatest artists of his generation.

Figure 2: Exit stage left… Nick Cave walks out of his first show in Chile in 1996. Source: YouTube
So, it was with interest that I picked up this article from The Guardian where somehow Nick Cave was involved in the whole ChatGPT conversation after one of his fans asked the algorithm to ‘write a song in the style of Nick Cave’ and sent the result over to the real Nick Cave for his opinion. Perhaps it’s best to quote the man himself to get a sense of his response:
Songs arise out of suffering, by which I mean they are predicated upon the complex, internal human struggle of creation and, well, as far as I know, algorithms don’t feel. Data doesn’t suffer. ChatGPT has no inner being, it has been nowhere, it has endured nothing, it has not had the audacity to reach beyond its limitations, and hence it doesn’t have the capacity for a shared transcendent experience, as it has no limitations from which to transcend. ChatGPT’s melancholy role is that it is destined to imitate and can never have an authentic human experience, no matter how devalued and inconsequential the human experience may in time become.
Clearly, he didn’t like it! The newspaper article picks up this fact but misses the key point in what he’s saying. The output of the algorithm has forced the artist to define what is unique about his art. In this case, it’s the human experience as the primordial magma where songs are forged. Nick Cave is hinting at something essential in the whole ChatGPT conversation, that predicting the next word in a sequence to generate credible text is not the same as writing a song. The melancholy role of ChatGPT is to imitate but never to experience. Perhaps ChatGPT could write a credible review of the Nick Cave concert I attended in 2002 in London but would not be able to express what that specific concert meant to me. ChatGPT would not understand the history of what happened before and the cultural baggage I carried with me to the venue that night.
If alternative rock gods are not your thing, take the example of education. Many schools are concerned about ChatGPT being used to cheat on essays and exams. Some districts in the US have banned it. Others are using this opportunity to question the kind of education we are giving our children where everything is geared towards exams, where teaching a child is like training an algorithm with some dataset and a goal. ChatGPT will produce a credible response without understanding the subject matter at all. This sounds like the educational experience I sometimes see with my children.
Schools are interested in how to spot text created with ChatGPT, but Sam Altman, the CEO of OpenAI, the company that built ChatGPT said this is not possible as the AI will adapt faster than our ability to catch it. In his own words:
Generative text is something we all need to adapt to. We adapted to calculators and changed what we tested for in math class, I imagine. This is a more extreme version of that, no doubt, but also the benefits of it are more extreme, as well.
More extreme indeed!
We probably felt at the time, that home schooling during Covid and the challenge of how to prevent cheating when kids had access to computers during exams would be the biggest challenge for a generation, but now that kids are back at school, that issue can be kicked to the long grass.
ChatGPT is presenting an even harder challenge. One that is more extreme and less predictable. Perhaps now is the right time to re-think the education approach and embrace a technology like ChatGPT that can help students access this vast source of examples, styles and intelligence. Maybe the point of education is for children to really learn and understand and to work with their AI buddies to put that knowledge to the test. Design ever more precise prompts to get the right answers, fact check the answers to make sure they are correct. Apply the best writing style following multiple templates suggested by the AI. In a world like this, humans can focus on understanding and bringing their unique experience and the AI can focus on predicting the next word in the sequence using an ever-larger knowledge base. I think it is more likely that we will continue to see educators pushing for ChatGPT to add watermarks to their technology so they can catch the cheats.
In my view, the most interesting contribution of ChatGPT is not in the large number of applications that people have dreamed up since it’s been available, but in the fact that it is forcing us to understand and define what it means to be human in partnership with these emerging artificial intelligences. This is a question that will affect all areas and will need diverse views and experiences to come together for a solution.
This is why I wanted to add my voice to the already saturated conversation about ChatGPT. I see this as an opportunity for something radical, however, I fear that we will see this as another case of a ‘faster horse’ and continue down the same track.
For now, we have people like Nick Cave and many others who will continue to define the human experience outside the Silicon Valley bubble. In my case, rather than writing a clickbait article citing the 10 things ChatGPT will allow us to do in 2023, I decided to reach into my personal story and remember a night in 1996 where I knew for sure that I would leave Chile one day. I thank ChatGPT for providing me with the prompt I needed to dig this out. Clearly the partnership is working. I just didn’t expect that it would be the algorithm prompting me rather than the other way around.
[1] This is done by collecting all the links originating from Reddit that point to other webpages on the internet and are rated as reliable by having more than 3 upvotes.
[2] The Guns and Roses concert sadly ended with a fatality after a crush at the entrance of the football stadium where the concert took place.

Arturo Dell


