I spent the last several days writing this. When I started, I felt like maybe the title was a bit of a hyperbole, even though I feel like my point is well argued. But then again, you have people working in this industry saying that they are building a weapon of mass destruction that could […]
Post
The term "AI" is like a miasma. It's hard to contain or define. For this particular exercise, I define through five properties. It has an agenda, which is filtered through a promise of what it can do, propelled forward by a collaborative political system, put in place by a specific business […]
Racism has been around for centuries, and the current iteration of "AI" for half a decade, so it seems foolish to compare the too. But specifically because of the rapid adoption of "AI" in the public sphere without criticality, some compare it to historical precedent like "the railroad" or "the […]
0 reposts0 likes
I look at how a technology like Eli Whitney's cotton gin had an effect on slavery in the United States. The tool, it can be argued, is not political. But in context, it is a destructive force (most historians point to the cotton gin as a major reason that the South became dependent on the […]
The difference with "AI" technology is that it doesn't just do one thing well. Even if it fulfills its job as "software engineer intern," there are so many problems with the tool itself (propaganda, spam, CSAM, mental health, etc...). Those problems should alone give pause. But in the wider […]
(previously broken the thread) https://hachyderm.io/@pythonbynight/117264348973344313
Local models or "ethically sourced" models as the solution is merely apologetics for the current "AI" landscape. They are mere drops in the bucket compared to what we're dealing with today. For example, Apertus, the open, GDPR compliant LLM developed in Switzerland boasts 70 billion parameters […]