What is our work for? – Poder Popular

J. Duarte | We work at Sigma AI and we are now facing a collective dismissal. If before this event the atmosphere in the company was already one of fear and repression, now we are confronted in a much cruder way with our own fragility and finitude. We have no way of answering, in our daily lives, to so much violence. In addition to their control over our bodies by counting the number of times we went to the toilet or went to drink water, our lives are now even more stripped of any rights when facing possible dismissal. The violence we suffer does not constitute a legal crime, it is even foreseen, and it makes us recognise our smallness, our fragility. It even seems that we should feel guilty for standing up to them. Even so, it is exactly at this moment that we feel that dreaming makes more sense, creating a line of flight from our daily life. We decided, me, Nina Gomes and Lucas F., to have some beers and talk about what our work really is and also what it could be. These are the reasons for this 3 way dialogue.

JD: Hi. After some time working at Sigma, I wanted to start this conversation by asking a very simple question: how are you feeling, day-to-day, in this workplace? What are the differences in relation to the period before the collective dismissal? I ask this because as I was suspended I really don’t know how the atmosphere is right now.

NG: We have been exchanging more because this story of the collective dismissal generates a «gossip» atmosphere, but the question of «Will we leave too?» is constant. We joke in the spirit of laughing with nervousness. One of the sources of my anguish is the lack of coherence and transparency in the company. If I don’t know where my work comes from and where it is going, I don’t know what I am doing or what for, apart from «training the machine». I don’t know what part of the cogwheel I am part of. The daily incoherence is unsettling and adds to the sense of uncertainty. The lack of response, the daily changes of plans, «now you can holiday, now you can’t, now you have to», etc.

LF: And there is a collective effort by the company to maintain this lack of logic; the answer to everything is: just keep doing it. Along with this, one can point out more practical day-to-day points, such as the time spent out of class; which in my opinion should be more flexible and coherent with the fulfilment of our work goals. For some people, the work done by the security guards is somehow uncomfortable, I have no personal problems with it, but neither the workers nor the security guards themselves understand very well why it’s being done. The head of the room won’t even let you in with paper, for adding another fact. And finally, lately with the new projects everything has become even more confusing and non-transparent. I don’t think any of us workers understand what we are doing.

JD: Right, the environment seems to be even more confusing then, not even better or worse but more confusing. I also always wondered what we were doing there, the whole issue of not being able to enter with our things in the workroom seemed to me a stage play of bad taste, with automatic characters and I also didn’t understand all that eagerness with the productivity targets… How is that going now, about the targets? Has the relationship between those who work there changed?

LF: Before we worked by objectives that were rewarded by bonuses for having been surpassed. The more we produced, the more we earned, to compensate for a very difficult and repetitive job. Since the beginning of 2023, the workload has changed drastically and new projects have come in, so the bonuses are no longer achievable. This created an even bigger limbo in the team, because producing more is no longer rewarded, in other words, we now have to comply with a lower workload, but following the same work schedule. Which in general creates an anxiety in the employees, because there is latency time. In this new scenario, just keep doing the same working routine we had before does not generate any bonuses to anyone. And the company could adapt in a certain way. The issue of bonuses, of the time inside the rooms…. in short, we have a way of working that no longer adapts to the work we are asked to develop.

NG: The bonus system is a way of keeping people in the dull job by the possibility of earning more than the base salary, you know, the minimum of what they can do. Because if you do more audios than the goal, the more audios, the higher your salary at the end of the month. I don’t think there should be a bonus. Either you create a target of number of audios or of working hours, i.e., now we have to comply with the schedule and also comply with a certain number of transcriptions. I would prefer to work fewer hours and no goals. We could, for example, have a goal of 2500 audios a day to be completed at a pace that each person finds more comfortable. If you finish in 5 hours, you get up and go.

JD: I really think the critical analysis of this particular company is done. We are rubbish, nothing but rubbish. But what about artificial intelligence, per se, what do you think of it? I honestly don’t think any technology is neutral but I have a really ambiguous feeling about the whole thing. At the same time that I can imagine other uses for voice recognition, what I was doing on a daily basis seemed really stultifying to me.

NG: I’ve been struggling to see the positive potentials of AI, I think we do a job training a technology that will soon replace us. My perception is that AI creates and will create a lot of low quality, dull, repetitive jobs with a very short horizon in time. Times will be tough. We’re fucked.

LF: Yes, I think it is useful, yes. Mainly because it can help people who don’t have the same access to technology. Looking at the case of Brazil, a country where many people still can’t read, the possibility of communicating orally creates greater access to technology, it makes this technology reach more people. The same can be applied to people with difficulties in the reading process, as is the case of elderly people who often have problems reading such small print on mobile phones or computers. It can also serve as a tool to speak other languages. And going further, we can think of the case of people who live alone. There are more and more people living alone and people are getting older, people live longer; if some problem or unforeseen event happened to a person living alone, the machine could find out about it; if the person speaks to the machine every day, the machine will find it strange if they don’t.

NG: There’s a system that is a necklace with a button connected to the telephone; in case of emergency, that button calls a central office that sends an ambulance if there is no response. However, if it’s a situation where the person is unconscious, that system no longer works.

JD: That’s totally true, in the pandemic a lot of elderly people died alone, in their own homes. Their already dead bodies were only discovered days later. I can also see a lot of potential in that technology, it is human knowledge after all. The problem is what Nina said: they could replace us, it would be possible to continue to create wealth without any intervention from a human body. At the same time, it’s a technology that can help children who can’t yet write or blind people who want to use a computer. I really feel a great ambiguity about this.

NG: If you develop an artificial intelligence to the point where it replaces the human, who can’t work more than 5 hours a day, the machine is going to work non-stop, we are going to reach the pinnacle of the exploitative system with an object that doesn’t demand, holiday, compensation, doesn’t get sick, but at the same time, that doesn’t mean that humans are going to have more free time, it just means that we will be left without sources of income.

LF: This model will have to change, because in a society where we live to work, when artificial intelligence replaces work, what will happen?

NG: We are going to live in misery! I’m a partisan of cyberpunk pessimism.

JD: Well, okay, technology is not neutral; and human life may even cease to exist still within capitalism. But can all this technology be re-appropriated or not? It’s human knowledge. It could be used in a society that is not centred on wage labour.

NG: We live in a system much bigger than ourselves that will always choose the violent, exploitative, segregating, excluding, racist side… How many things were invented with the intention of advancing and ended up being used for those ends? I think this is our paradigm, unfortunately. Could we do it? Obviously it could be done, but I can’t believe in that, I don’t think it will happen. But okay, if I allow myself to dream, it would be amazing to imagine that one day the work would be to train this AI to actually incorporate accents, to include street vocabulary, technology to reinvent the education system (which urgently needs a reform), that would be beautiful. Like making internet more accessible for people with disabilities, people on the autistic spectrum, it would be more rewarding to have those perspectives in mind while working.

LF: This accent thing is very curious because we had already noticed with the transcriptions, that the machine has difficulties understanding certain accents; in Brazil, for example, the accent in the Northeast region is quite misinterpreted by the artificial intelligence. The machine does not understand the accent of the Northeast and that would be an interesting job.

NG: Being stimulated to skip audios that we don’t understand (because we only listen to them twice) leads us to discard speaches with accents and languages from specific regions, speaches that if we could dedicate extra minutes would allow us to understand them and feed the AI with more nuances of the language, trying (at least) to go against the homogenising tide of the big media.

JD: Yes, our daily lives could be very different. Making fewer skips and trying to integrate more diverse audios into the machine could be just the beginning. Would it be possible to subvert today’s technology? Because, of course, everything could be even worse, you know… What seems clear to me is this warning, a fire alarm: either we start dreaming more or this could end very badly!

Entrevista en castellano: