Eat my ass OP. I need the help.
Damn right.
I can finally dictate my work and have it come out with the right words, the right spelling and the right punctuation.
Rather than spending longer cleaning up the dictations than I saved by not typing.
F**k AI for its nonsense and hype, but glad to have a better spell checker.
To be fair “fuck AI” is about generative AI specifically instead of spel-checkers and weather predictors.
For example the “AI image generators let disabled people do art now” argument.
Or even, “fuck your privilege “! Unless my wife wants to paint holding the brush in her mouth don’t kick her in the expression. You want royalties for artist or advance optin for model training I’m with you.
Edit:Didn’t realize the community I was in. Sorry all.
AI (ML) transcription and image/text recognition is helpful, but that came out so long before the slop generators that I’m not sure it can be put under the same ‘AI’ umbrella.
Its being used for language translation too, right? That actually feels like a useful thing
This is my primary point of hate on genai: they are barely more ai than flipping Markov Chains or Horse_ebooks.
And OP speaks for all disabled people or something?
I am pretty sure the kind of AIs that generally help with accessibility are not the same as an LLM or other generative type AIs. They’re statistical models and shit like that, and have existed much longer.
This is the problem of using an already not-quite-true term (AI) for several wildly different things.
not-quite-true
Hell, not at all true. Nothing we currently have is intelligent, they are all complex statistical models that are described using misleading terms in the interest in getting the public to believe they are more capable than they are. Reasoning, learning, & hallucinations are all hyped up to make them sound intentional instead of statistical.
Hell, not at all true
To be perfectly fair: the “artificial” part is true. 😌
Artificial requires intelligence to be true to have any meaning. Yeah, take that!
(Also you were right to point that out.)
Intelligence doesn’t mean sentience. Insects can perform intelligent behavior. There’s nothing wrong with the term AI. It’s people confusing the word to mean Data from Star Trek.
It’s people confusing the word to mean Data from Star Trek.
People like those promoting AI as replacements for workers?
A computer used to be the title of a job a person performed. Electronic computers displaced millions of human computers out of their job. LLM’s are doing a similar thing. It doesn’t need to be sentient to provide a small efficiency boost that results in fewer employees overall.
Human calculators were doing repetitive math, some at a high level, but the majority were literally doing a large amount of basic math. The ones who were doing the high level stuff didn’t lose their jobs, they benefitted from the automation, like with manufacturing automation and computer algorithms that have structured and repetitive outputs.
AI is being shoved into positions where some level of flexibility in clarifying context and decision making is necessary. It is currently being used to impersonate therapists, trample creative jobs by scraping copywrited works, and forcing itself into everything it can so it can scrape more and more data. It isn’t comparable at all.
I work disability support and have found ChatGPT to be really helpful for someone I work with who has cerebral palsy. Basically he prompts GPT to make his sentences more concise and it does just that.
For him, that’s really helpful.
Of course, to each their own, but I believe it does have its place in assisting people with disabilities IF THEY WANT AI ASSISTANCE
I teach a course in disability in education settings and about 10% of learning disabilities are a writing disorder called dysgraphia. You can think of it like a writing version of dyslexia (which is like, 80% of learning disabilities). LLMs are absolutely a reasonable accomodation in this case as well because it can help people who have trouble putting their thoughts into words.
That’s just one example, and generally what I teach is scaffolding – that is, you try not to over rely on tools and use them when needed to improve learning when learning would otherwise not be possible. After all, dyslexic people still need to be able to read on occasion, and dysgrapic people need to be able to write, and so on.
(And do keep on mind I teach and LLMs are a bit of a nightmare in grading so do understand I say this use case still very much not a pro-AI person, lol)
- Learn how to talk with a person
- Make them rely on a service owned by a billionaire
Everything is owned by billionaires. The health care system is owned by billionaires. Are you saying people should refuse health care that improves their lives for that reason?
Fix the system. Don’t shit on people trying to survive it.
I don’t understand this guy. What did he say?
He’s responding to a post saying AI use for accessibility purposes is probably the only moral use.
I mean, fuck AI. But if it helps accessibility and exists, I’m in favor of people using it for that if they choose.
