melica.me
🇨🇦 | Human-centric ML 🐈🐈⬛🎥 📕 🎮 🧶 📸 | This is a personal account all views are my own |
163 posts
126 followers
206 following
Regular Contributor
Active Commenter
comment in response to
post
Many researchers in AI communities similarly fear that the research they work hard on a d freely publish will be used by AI giants without credit or consideration for the true goals behind said research. Many feel discouraged to continue to share their research freely
comment in response to
post
All this aside from the fact that water consumption is not a good indicator on its own for the sh*** environmental impacts of these models
comment in response to
post
That hamburger chart going around on Chat GPT water consumption is such BS and it’s come back around so much I get 🤢 looking at it. SA quoting the buger chart should be another proof for the heavy PR campaign that has little to do with the truth, and the willingness to double down on the lies
comment in response to
post
Probably gonna leave behind the people who collect the little people’s taxes.
comment in response to
post
They probably have devs name them, next model is gonna be “o3-preview-final-2”
comment in response to
post
Article says it’ll do deep reasoning as well but only if you want it to and you’ll be able to control the amount of reasoning it’ll do.
comment in response to
post
Thanks for sharing! Curious what you mean by “performance”. performance on benchmarks? Inference speed?
comment in response to
post
Her last name is FreeLAND not FreeMAN 🤦♀️😆 wtf so ironic
comment in response to
post
For this case it was already acknowledged that the other company had used the dataset/IP. A really big challenge in other cases is to get them to admit or prove that they did use your data in the first place, before it even gets to fair use…
comment in response to
post
👏👏👏
comment in response to
post
Not sure if this has been your experience as well but now more than ever I sense a push to make women in tech invisible. Like we’re going backwards. People giving women a voice to then use it for a different agenda. But also, I really wish they’d stop making THIS guy the face of AI. 🤦♀️
comment in response to
post
Yay for the chaos hair. The right age for it ✊😅
comment in response to
post
Will you by any chance be writing a blog/article about your experience there? Or know others who will write such a thing I can look out for?
comment in response to
post
Glad this was made public! Thank you :)
comment in response to
post
Thanks! Would be curious to see the trend for tech occupations within non-tech companies
comment in response to
post
Is this jobs in tech companies? Or jobs with a tech nature regardless of the company? Like software engineers in a health company count or not? I heard that there is a trend toward incorporating more tech roles in non-tech companies, so maybe that’s the unseen side of the trend you showed?
comment in response to
post
Agreed. GenAI pandora’s box is open. Can’t be closed. Research like this is useful to help navigate how and which goals should be incorporated into these models and who should get a say
comment in response to
post
I feel personally attacked with this 😆
comment in response to
post
Of course!
comment in response to
post
Yep they often take these as suggestions rather than cautionary tales
comment in response to
post
Also this came out recently:
arxiv.org/abs/2501.17420
Which bolsters your point IMO
comment in response to
post
Gladiator fight of AI ads
comment in response to
post
Much of my grad school research went toward studying what makes women interested in STEM and how STEM pathways could become more attractive to women :(
comment in response to
post
Awesome!! 💃🏻💃🏻💃🏻
comment in response to
post
Google just retracted their responsible AI principles that said they won’t create things that hurt humans or they won’t make weapons. Why do you think they did that? They don’t have to pretend that they want “good” AI anymore. They’re straight up going for the weapons.
comment in response to
post
They can be dangerous! Regardless of how powerful they are at the moment they have the potential to be very dangerous. Their power will grow, even if slowly, and with that growth their danger will grow. But monopolies should not hoard it and allow development and safety research by other parties.
comment in response to
post
lol ofc.