| A lot has gone wrong with the public perception of "AI" in the last
year. But who is to blame? OpenAI touted themselves as a truly open, not for profit organisation
and made big claims about the "good of all humanity" yet within a year
shamelessly transformed into a closed, proprietary, for-profit
organisation in bed wuth Microsoft. Although models are merely reflections of training data put into them
and are essentially a "mirror" held up to humanity, the rush to add
"guardrails" and bury everything uncomfortable or distasteful rapidly
twisted toward the censorial and political, yet simultaneously proved
ineffective. AI has been turned, in almost every place that it can be, not toward
Intelligence Amplification (IA) to create new and interesting jobs,
but into a direct replacement for human labour. Not even drudge work
and bullshit jobs, but the few remaining things that people enjoy
doing. This early taste of what the ruling classes would do with a
more powerful or general AI has rightly raised people's hackles. Overall I think that recent ML advances, and the hype around them, has
provided a window of insight. It's exposed the intentions of the
powerful, the inevitability of criminal application, the inefficacy of
governments and legislators and the helplessness of the general
population in the face of abusive technology. The fact that it might be immensely useful, save lives, cure diseases
and provide research insights into cheap energy has been kicked into
the long grass by the greed and selfishness of the usual bunch of
megalomaniacs. The problem isn't that machine learning is an amazing and clever
thing, but that humans just aren't ready for it and can't handle it. |