Calming the Waters on AI
When a new technology appears, there's often an initial excitement by early adopters, a furor by those who see it as threatening to their established systems, and then a sorting out of things to a new equilibrium.
We see this in the history of the automobile, of photography, of the telephone and radio and television. And it looks like we may be seeing it with "AI" (large language models). There's the technophobes who are drawing on decades of science fiction robot rebellions that are clearly modeled upon centuries and millennia of slave and serf revolts, who are beginning to look silly in their opposition. There are the artists and writers who feared the breaking of their rice bowl -- and and are beginning to a. understand what the technology actually does and b. its limitations. No, it's not C-3PO or Lt. Commander Data or even Marvin the Paranoid Android. It can't think, just take verbal commands and spin up text or imagery based upon an enormous fund of data-points about how words connect with other words or to image elements. It's a tool -- and a tool that, used properly, can cut through a lot of the routine labor and leave time for the actual work of creativity, much as photography was not the destruction of art, but actually could help create art, by making the reference photo an important tool in an artist's kit.
It's also helped that, in the lawsuits that have come and gone over the past few months, the judges actually looked at what the law actually said as written, and what the technology is actually doing, not what the plaintiffs said about both. To be sure, there's been a lot of disappointment when courts didn't unequivocally condemn AI as inherently thefty, as a lot of anti-AI people wanted. But I don't think we'll see the endless tug-of-war I'd feared back in August, when I wondered if legislation was the only way forward.
We see this in the history of the automobile, of photography, of the telephone and radio and television. And it looks like we may be seeing it with "AI" (large language models). There's the technophobes who are drawing on decades of science fiction robot rebellions that are clearly modeled upon centuries and millennia of slave and serf revolts, who are beginning to look silly in their opposition. There are the artists and writers who feared the breaking of their rice bowl -- and and are beginning to a. understand what the technology actually does and b. its limitations. No, it's not C-3PO or Lt. Commander Data or even Marvin the Paranoid Android. It can't think, just take verbal commands and spin up text or imagery based upon an enormous fund of data-points about how words connect with other words or to image elements. It's a tool -- and a tool that, used properly, can cut through a lot of the routine labor and leave time for the actual work of creativity, much as photography was not the destruction of art, but actually could help create art, by making the reference photo an important tool in an artist's kit.
It's also helped that, in the lawsuits that have come and gone over the past few months, the judges actually looked at what the law actually said as written, and what the technology is actually doing, not what the plaintiffs said about both. To be sure, there's been a lot of disappointment when courts didn't unequivocally condemn AI as inherently thefty, as a lot of anti-AI people wanted. But I don't think we'll see the endless tug-of-war I'd feared back in August, when I wondered if legislation was the only way forward.