It's a mistake to think that AI ethics means to bet against AI. It's betting that AI can be a force, if properly used, to help people thrive and flourish. AI ethics is very much present in the movement for "Human-Centered AI" and in the efforts behind a "Humane Technology". Any one who thinks that AI should serve to augment, enhance, and improve human agency is an AI ethicist at the core.
Oh, I agree. I have nothing against AI ethics in principle.
But one thing is what AI ethics should be and another what AI ethicists (not all, just some of the most vocal among then) are doing - or better, how they're doing it.
Totally agree. I also think that "algorethics" is a distraction because ethics is not about algorithms or codes, but about humans who write codes/algorithms which affect themselves and others. AI ethics cannot be reduced to algorithms no less than ethics reduced to rules and laws. You cannot legislate or codify everything. AI ethics is not about compute or data either. Ethics is all about human decision-making and agency, in this case, in human-computer interaction.
Your logic is hard to resist. It is unfortunate the way the AI ethics hand was played. I personally think the “everyone’s AI truth” is an important driving factor in the diminishing of the spirit of the cause, but as you highlight, the primary driver is that these companies serve as the primary locus of innovation and thus power.
I don't think they made any mistakes. I am genuinely surprised you put it like this.
The question about ethics is not a rhetorical one. If something is ethically dubious, it doesn't become less so if the person pointing this out is rhetorically less than convincing.
Nobody would ever listen to ethicists in the climate.
They were like people talking about the crusades and the inquisition at the Vatican.
People at the Vatican know those things are an issue but to them, because they believe, they're more like a technical problem to be fixed, not a destructive stain on the Catholic project.
And now people can make a fortune off AI, so are they really going to listen to ethicists?
This was never going to work. It's like trying to get the finance industry to listen to ethicists. As if!
Some people only pretended to care about ethics as they were building their personal brands as deep thinkers coming out of universities where people take ethics seriously.
I say that much in the article. It was always an uphill battle.
Also yes, they made mistakes and still do. Not all of them, not all the time, but using dismissive metaphors to criticize the technology as a technology is a mistake and the consequences are that they lose the spotlight.
This isn't an article to criticize AI ethicists. It's an analysis of where they lost a battle that was already very hard in a way that I think could've been avoided.
(Their fight being inherently hard was actually a factor in reaching the mainstream, as it happened to Gebru and Mitchell when they were fired.)
I work in academia and the nonprofit sector. I feel that AI Ethics as a concentration or field is alive, well, and growing - especially when it comes to academic research and educational initiatives for the nonprofit sector and policymakers. People are starting to see how AI technologies can transform nearly every field. I see it called by many names, like "Responsible AI," or "Human-centered AI," but I think there's a growing interdisciplinary and cross-sector movement to examine both the benefits and risks of AI technologies, work to mitigate and prevent harms and biases, and ensure that human - AI partnerships are human-driven and beneficial to society. Many "ethicists" I interact with are not trying to be negative or alarmist - at the heart of it, people are trying to leverage AI for social good and educate and empower people to take ownership over how they use or don't use technology and manage their private data. Sometimes people are just lacking communication skills, and I think communicators have a significant role to play here.
It's a mistake to think that AI ethics means to bet against AI. It's betting that AI can be a force, if properly used, to help people thrive and flourish. AI ethics is very much present in the movement for "Human-Centered AI" and in the efforts behind a "Humane Technology". Any one who thinks that AI should serve to augment, enhance, and improve human agency is an AI ethicist at the core.
Oh, I agree. I have nothing against AI ethics in principle.
But one thing is what AI ethics should be and another what AI ethicists (not all, just some of the most vocal among then) are doing - or better, how they're doing it.
Totally agree. I also think that "algorethics" is a distraction because ethics is not about algorithms or codes, but about humans who write codes/algorithms which affect themselves and others. AI ethics cannot be reduced to algorithms no less than ethics reduced to rules and laws. You cannot legislate or codify everything. AI ethics is not about compute or data either. Ethics is all about human decision-making and agency, in this case, in human-computer interaction.
Your logic is hard to resist. It is unfortunate the way the AI ethics hand was played. I personally think the “everyone’s AI truth” is an important driving factor in the diminishing of the spirit of the cause, but as you highlight, the primary driver is that these companies serve as the primary locus of innovation and thus power.
I don't think they made any mistakes. I am genuinely surprised you put it like this.
The question about ethics is not a rhetorical one. If something is ethically dubious, it doesn't become less so if the person pointing this out is rhetorically less than convincing.
Nobody would ever listen to ethicists in the climate.
They were like people talking about the crusades and the inquisition at the Vatican.
People at the Vatican know those things are an issue but to them, because they believe, they're more like a technical problem to be fixed, not a destructive stain on the Catholic project.
And now people can make a fortune off AI, so are they really going to listen to ethicists?
This was never going to work. It's like trying to get the finance industry to listen to ethicists. As if!
Some people only pretended to care about ethics as they were building their personal brands as deep thinkers coming out of universities where people take ethics seriously.
I say that much in the article. It was always an uphill battle.
Also yes, they made mistakes and still do. Not all of them, not all the time, but using dismissive metaphors to criticize the technology as a technology is a mistake and the consequences are that they lose the spotlight.
This isn't an article to criticize AI ethicists. It's an analysis of where they lost a battle that was already very hard in a way that I think could've been avoided.
(Their fight being inherently hard was actually a factor in reaching the mainstream, as it happened to Gebru and Mitchell when they were fired.)
Yeah, you are probably right.
It's very plausible that is the issue. Nobody will listen to you at that point.
The problem is that they probably wouldn't have anyway.
The hype was a shiny transient thing. It wasn't going to last. They should have kept focus.
I work in academia and the nonprofit sector. I feel that AI Ethics as a concentration or field is alive, well, and growing - especially when it comes to academic research and educational initiatives for the nonprofit sector and policymakers. People are starting to see how AI technologies can transform nearly every field. I see it called by many names, like "Responsible AI," or "Human-centered AI," but I think there's a growing interdisciplinary and cross-sector movement to examine both the benefits and risks of AI technologies, work to mitigate and prevent harms and biases, and ensure that human - AI partnerships are human-driven and beneficial to society. Many "ethicists" I interact with are not trying to be negative or alarmist - at the heart of it, people are trying to leverage AI for social good and educate and empower people to take ownership over how they use or don't use technology and manage their private data. Sometimes people are just lacking communication skills, and I think communicators have a significant role to play here.