ChatGPT, I have already said it on several occasions, seems to me to be an absolutely disruptive technological innovation. For how well it works, for how its creators have been able to make it articulate responses within the reach of all audiences, for its ability to hold conversations, for its versatility… in short, the list of praise that I believe this OpenAI artificial intelligence deserves it is long and classy. And yes, I know that he makes mistakes, but we must bear in mind that his potential for improvement is overwhelming.
Now, not everything is positive around the fashionable technological service. It has a much darker face and no, unfortunately I am not referring to the errors that I mentioned before, and neither are I referring to the apps that try to monetize a free service in their own interest. No, I’m talking about a business decision that can be legal, of course, but that is ethically so reprehensible that it can detract from a service that would otherwise be worthy of glory.
Relocation, relocation or whatever we want to call the technique of subcontracting certain services in countries where it is much cheaper is not something new. The Anglo-Saxon world has been doing this for decades, mainly in India, and as for our country, calls to many customer service services remind us that most of these calls are answered from call centers in Latin America, for speaking in both cases only of a specific type of activity.
Mind you, I’m not saying this is bad. per se, because in certain circumstances it is true that it can be a great stimulus for those local economies, and when that happens in developing countries, it becomes a factor that substantially boosts them. In addition, the difference in wages can improve the competitiveness of companies, something that we like or not, today is part of the business creed that they must subscribe to to survive.
The problem, which is questionable, is that companies from the so-called first world resort to offshoring to pay miserable wages even in the battered economies of the countries chosen for this purpose. And according to an investigation carried out by the prestigious Time magazine, we are faced with a case of this type, since OpenAI hired Kenyan workers for less than $2 an hour for ChatGPT.
The investigation, citing multiple sources, reveals that the take-home pay of outsourced workers in Kenya, through Sama, a San Francisco-based company that employs workers in Kenya, Uganda and India and is used by many tech companies, It ranged from $1.32 to $2 an hour, depending on their seniority and performance. Figures that should make those responsible for OpenAI blush, even more so in these times when multimillion-dollar investments are coming in the company thanks to ChatGPT.
Workers were outsourced for a very important part of ChatGPT, feeding the AI toxic language, hate speech and such content so that the AI is able to detect it and therefore remove it. After the precedents with other AI, many of the first tests carried out by the general public with ChatGPT have consisted of checking if it reflected some type of bias in this regard, but the system seems to be very well educated in this regard.
The evolution of artificial intelligence is amazing, but in cases like ChatGPT it does so at the cost of job insecurity, according to the Partnership on AI, a coalition of AI organizations of which, curiously, OpenAI is a part: “Despite the critical role these data enrichment professionals play, a growing body of research reveals the precarious working conditions these workers face.”.
It is curious, because we have been talking about ethics in artificial intelligence for quite some time now, a very interesting debate that still has a lot to say, but it is paradoxical that we focus on the ethical behavior of artificial intelligence, while the development of them is based, at least partially, on highly questionable practices on an ethical level. Legal, yes, but highly questionable. Perhaps, even if it is only in part, we should learn a bit of that ethics that we want to teach to machines because, otherwise, it is possible that with the passage of time they will also advance us in that sense.
