I usually enjoy Dr. Fatima's well researched videos and I did like this one too. It made me ponder my own use; why do I use it instead of asking a person for advice on e.g workout. A lot of it...
I usually enjoy Dr. Fatima's well researched videos and I did like this one too. It made me ponder my own use; why do I use it instead of asking a person for advice on e.g workout. A lot of it boils down to convenience:
I don't know a lot of people with expertise in the field.
I want an answer to my question now.
I mostly don't want to pay for the advice.
If I didn't use an LLM I'd probably be asking the same questions, but I'd be using a search engine and trawl through forum posts instead.
Anyway, the video discusses the harms of generative AI focusing on LLMs. It discusses what effective ways are of reducing those harms and calls out shaming as a bad way of reducing others use. The last point is kind of central and worth reflecting on (not just in the case of AI). It ends on suggesting a few alternate avenues that is likely superior in reducing the actual harm caused by AI usage in society.
Aside: I always have a hard time picking n area where to post commentary on tech. I mean this isn't really tech, it is more society related in general. Feel free to move it :)
Oh, didn't expect a Dr Fatima video to pop up here, I've enjoyed her content for a while starting with colonialism in astronomy. I just watched this one and shared it with my friends yesterday....
Oh, didn't expect a Dr Fatima video to pop up here, I've enjoyed her content for a while starting with colonialism in astronomy. I just watched this one and shared it with my friends yesterday. They're the sort of hard line shamers she talks about, I don't expect any to click the video. It's disappointing to see them talk and act the way they have been when I know they're capable of the more empathetic and effective approach described by Dr Fatima. It's frustrating to watch folks I care about get so upset over a topic they refuse to learn about. Their anxiety could be assuaged if they'd engage, learn how to effectively resist and reduce harm. They feel the same AI anxiety she describes so I'm patient; I can only lead a horse to water, not make it drink.
I've tried other tactics, like when I recently tested out the rnj-1 model, I gave them a write-up on its capabilities, cost to run, estimates on cost to train, company background and funding, that sort of thing. I got some engagement there and was able to communicate some of the nuances described here.
I usually enjoy Dr. Fatima's well researched videos and I did like this one too. It made me ponder my own use; why do I use it instead of asking a person for advice on e.g workout. A lot of it boils down to convenience:
If I didn't use an LLM I'd probably be asking the same questions, but I'd be using a search engine and trawl through forum posts instead.
Anyway, the video discusses the harms of generative AI focusing on LLMs. It discusses what effective ways are of reducing those harms and calls out shaming as a bad way of reducing others use. The last point is kind of central and worth reflecting on (not just in the case of AI). It ends on suggesting a few alternate avenues that is likely superior in reducing the actual harm caused by AI usage in society.
Aside: I always have a hard time picking n area where to post commentary on tech. I mean this isn't really tech, it is more society related in general. Feel free to move it :)
Oh, didn't expect a Dr Fatima video to pop up here, I've enjoyed her content for a while starting with colonialism in astronomy. I just watched this one and shared it with my friends yesterday. They're the sort of hard line shamers she talks about, I don't expect any to click the video. It's disappointing to see them talk and act the way they have been when I know they're capable of the more empathetic and effective approach described by Dr Fatima. It's frustrating to watch folks I care about get so upset over a topic they refuse to learn about. Their anxiety could be assuaged if they'd engage, learn how to effectively resist and reduce harm. They feel the same AI anxiety she describes so I'm patient; I can only lead a horse to water, not make it drink.
I've tried other tactics, like when I recently tested out the rnj-1 model, I gave them a write-up on its capabilities, cost to run, estimates on cost to train, company background and funding, that sort of thing. I got some engagement there and was able to communicate some of the nuances described here.