Recent searches
Search options
As someone who has been doing "AI" professionally for 10 years now and really believes in its potential, I've been quite bummed about how much people on Fedi hate it.
But I've just set foot outside academia for the first time and heard a keynote talk about AI in an industry meeting and it was so fucking dumb that I left early, angrily. The most insane brain rot bullshit I ever came across.
Now I understand why people outside academia hate it so much.
#AI #ML #MachineLearning
@maltimore Mostly, the people I see here don't hate machine learning. They detest the way hardly-helpful, intrusive and privacy shattering "products" are forcefully introduced into their lives with little recourse. There's a difference between the inherent potential of a tool to be useful and its manifested application for the user.
@fabianegli @maltimore Yes, it seems that all the low hanging fruits are mostly annoying or privacy invasive while the truly useful stuff takes more effort and thoughtfulness to employ. However, the general interest is also helping the development of the good stuff, so I guess we’ll just have to live with the slop.
In #archaeology we love *analytical* machine learning applications, like the ones that can search through aerial photographs for something we're interested in, or make a charred scroll legible.
But we mostly hate *generative* applications, like LLMs and prompt-based image generators. Because they debase scientific knowledge with their hallucinations, and undermine skilled writing and painting.
@mrundkvist @fabianegli @maltimore
As a literary scholar this is me too. I love the chance to read incomprehensible manuscripts. I hate that co-pilot always tries to predict my next sentence. I really want and NEED to think that sentence myself.
@mrundkvist @fabianegli @maltimore
Yes, I have found a way to turn it off in word, I will find it in outlook and then when the US goes pear shaped I will find a way to turn it off in the new european programs as well.
@mrundkvist @fabianegli @maltimore this is the key difference in #ai, *analytical* machine learning used by scientists and *generative* machine learning, used by charlatans.
I believe the second one is also known as Grand Theft Autocorrect.
@mrundkvist @fabianegli @maltimore Jumping in sideways here, but if a software engineer wanted to look for work doing this sort of thing, what would be a good way to search? I'm not familiar with the terms archeologists use to describe these roles.
It would be in #academia. You would either be a computer science lecturer with research time built into your job, or get hired onto a funded project headed by an archaeology professor.
@mrundkvist :grins: adding this to my list of potential futures I could be really into and should keep an eye out for openings in, thank you!
@feonixrift Wishing you the best of luck!
@mrundkvist @fabianegli @maltimore I think this distinction deserves a much wider audience
I only just recently learned the technical terms for a distinction that is obvious to people who have used these tools. There is nothing obvious in common between an analytical tool that tells me where the charcoal burning sites are in a big lidar dataset and generative tools such as #ChatGPT or the image generators.
@fabianegli @maltimore I'd suggest thinking people do hate machine learning deliberately divorced from humanity. If humans were hired full time to monitor, regulate and check, such people would be on board. So, *Paying For* and *Respecting" People/Humans.
= Exactly what tech bros DON'T want.
@fabianegli @maltimore agreed, as with any tool.