Profile avatar
ontologic.design
enhancing mankind’s ability to access and harness information is a public good programmer // poet // poster
221 posts 96 followers 79 following
Regular Contributor
Active Commenter
comment in response to post
Online socialists is exactly who I’m talking about lol.
comment in response to post
You ppl slurp down republican propaganda like it’s air
comment in response to post
One of the few things the tumblr left got absolutely right tbh. It’s a shame the ethos behind it didn’t get spread
comment in response to post
In the bigger picture, the attitude of “it’s not my job to educate you” has largely won out, and it leaves ppl seeking right wing content bc those influencers are more than happy to “educate”
comment in response to post
That’s one piece of it. I don’t blame ppl leaving at all tbh bc it sucks there now, but if one is serious about trying to shift American politics back left I think you have to grit your teeth and fight through it. It’s still the highest leverage platform
comment in response to post
Are you illiterate, or just desperate for a dunk? The exact question this thread has raised is what the responsibilities of dataset curators are in regards to the terms of service. Pls stay on topic.
comment in response to post
*circle to square
comment in response to post
It’s the exact same thing as other public platforms. Posting on bluesky does not make you a service provider even though you have followers etc.
comment in response to post
There’s no square to circle. You are misrepresenting and misunderstanding what’s happening. A Huggingface employee sourcing and publishing a dataset on his own is not a service provider. Huggingface itself is a service provider. This distinction is easy to get
comment in response to post
I’m using it to simulate the minds of anti-AI ppl and then I’m gaslighting them all into loving it. You were the first to fall actually
comment in response to post
It’s not. A service would imply an ongoing commitment to performing action(s) on behalf of someone else. A dataset is a one and done drop. Huggingface is a service provider, individual datasets hosted there are not.
comment in response to post
No it’s not lol. It’s a dataset.
comment in response to post
If you didn’t want your posts publicly available you should not have posted to a service that is a public and open source protocol. ¯\_(ツ)_/¯
comment in response to post
Also you didn’t even read the TOS properly 😂😭
comment in response to post
Enforce it lmao
comment in response to post
Thank you for this work! Freedom of information is important
comment in response to post
Yeah you should just be smarter tbh
comment in response to post
This explains some of the holes in my twitter csvs
comment in response to post
I’m simulating your brain on my computer and I’m putting it through the trolley problem over and over. Do u want to know what your choices have been?
comment in response to post
Bold and controversial of me to say I know
comment in response to post
Indeed. Lmk if you have any ideas
comment in response to post
I intend to purchase it as soon as I actually have money 🫡
comment in response to post
Microsoft trained a state of the art text embedding model off of purely synthetic data arxiv.org/pdf/2401.003...