On 'Using ChatGPT 50-70 Times a Day' and the Philosophy of the Consulting of Oracles

Anyone wishing to catch a glimpse into the future would be well advised to read this short article from Business Insider, in which Conor Grennan, Dean of MBA Students at the NYU Stern School of Business and self-confessed ‘frequent ChatGPT user’, gives us his pearls of wisdom about how to use the programme to ‘level up [one’s] personal and professional life across any industry’. In my case - I don’t know about you - these are pearls cast before luddite swine. Bear with me while I grunt out a few observations.

First, we have arrived at the strange position that it is now possible for serious people (a Dean of MBA Students at the NYU Stern School of Business is presumably in that category) to talk about ‘levelling up’ in contexts that are definitely not ludic in nature. That is to say, any human being who is used to interacting with other human beings in a sincere and honest way - that is, who is used to thinking about other human beings as ends in themselves and not merely means to ends - knows instinctively that it is morally questionable to view relations with them as a game and to conceptualise ‘success’ on those grounds. Yet here we have somebody who ex hypothesi is a serious person (and clearly by any metric successful) informing us that not only is one’s ‘personal and professional life’ implicitly like a game, but also that in order to succeed in that game we should be using AI to generate, in advance, bespoke ‘professional talking points’ to use in conversations with individual people that we meet, as well as to inform us of those people’s personal interests so we can ‘bond’ with them. In other words, before meeting anybody, we should have used AI to find out what we can about them and thought about how we can use that to our advantage in forming a relationship with them.

I don’t know what word to use to describe that approach other than ‘sinister’. It reduces other people to something akin to computer sprites for us to manipulate, and our interactions with them to gamifiable set of processes and procedures. Social media can already, if you squint at it, be seen as encouraging somewhat sociopathic tendencies in heavy users, eating away at our capacity to appreciate that human beings are more than the sum of their parts (a photo, some posts, some personal data). This approach would put that on steroids if widely adopted - and it says a great deal about how artificial and inhumane modern professional life has become that anybody would suggest it is an appropriate way to behave. Business networking, especially among fancy-pants people, has always been phoney and instrumentalised. But that’s surely preferable to the incipient psychopathy that would result from the widely deployed meta-phoniness of ‘bonding’ with people based on secretly AI-generated profiling.

Second, it is fascinating that so many of Grennan’s recommendations are designed to plaster over, with technology, problems (I’m tempted to call them pathologies) that in themselves arise from an over-reliance on technology. The two examples that leap to mind are using ChatGPT to help tailor job applications to HR algorithms, and using ChatGPT to summarise long email chains to figure out whether they contain any ‘action items’ relevant to oneself. I have no doubt doing those things could be useful. But shouldn’t we be a little concerned that academics at one of the most prestigious business schools in the world are not asking themselves deeper questions about why it is that HR departments have become so used to making hiring decisions in the way that they do; why it is that so many of us have jobs in which we are esentially managed via email; and whether these are necessary features of our working lives, or just unnatural and unproductive cultural practices that we adopt due to path dependency, and would be best jettisoning? Perhaps it’s possible to solve the problems associated with overreliance on email, for instance, by using email less and having more face to face conversations and phone calls? And maybe that would have the corollary of making work nicer, more personal, and less stressful?

Third, and most importantly, what is astonishing to me is that Grennan does not appear to have reflected on the fact that there isn’t a great deal of difference between what he is advocating and the way in which his ancestors, thousands of years ago, presumably consulted soothsayers, shamans, ‘the gods’, and other oracles when trying to decide what to do and how to behave in difficult circumstances:

If you want to write a business plan, that's going to take you a week — it's a lot of brainstorming, talking to a lot of people, and market research. ChatGPT will do it in 30 minutes.

-is not in other words all that different to ‘If you want to go to war against the Saxinfrax people, it’s a lot of preparation, a lot of training, a lot of planning. The witch doctor will tell you what to do and whether you will win after just a short ritual.’ It is, in short, simply a way of delegating a difficult task that one finds inconvenient and bothersome. There is nothing wrong with that in itself, but there is something wrong with convincing oneself that the external force to which you are delegating it is going to be better at doing the work than you are, when you are the one with all the tacit and embedded knowledge and the incentives to perofrm well.

Moreover, we all know that the more reliance one puts on external forces to do things on one’s behalf, the less incentive there is to develop one’s own capacities to perform those tasks. The story of humanity’s interaction with technology is of course one of increased reliance and concomitant loss of skill; now we have central heating many of us are now incapable of building and starting fires, for example. But the fact that this pattern endlessly repeats itself and there are many instances in which the trade-off can be said to have been worthwhile does not mean that in each and every circumstance that is true. On balance, it is probably better that we have central heating than having to rely on our wits to generate heat each day. But is it better to rely on ChatGPT to tell us how to interact with other human beings, learn new things, apply for jobs, get superglue off our fingers, and read long chains of emails than it is to learn how to do those things better for ourselves? I am not convinced. And I am worried that purportedly serious people do not even seem to think it worthwhile asking such questions.

Now, back to my mud pit.