Fiona Melzer (@FionaMelzer) header
Fiona Melzer (@FionaMelzer) profile picture

AI Ethics | Cognitive Science | DEI | Feminism | Climate Change (She/Her)

Joined March 2021
Photo by Fiona Melzer (@FionaMelzer)Photo by Fiona Melzer (@FionaMelzer)Photo by Fiona Melzer (@FionaMelzer)Photo by Fiona Melzer (@FionaMelzer)Photo by Fiona Melzer (@FionaMelzer)Photo by Fiona Melzer (@FionaMelzer)Photo by Fiona Melzer (@FionaMelzer)Photo by Fiona Melzer (@FionaMelzer)Photo by Fiona Melzer (@FionaMelzer)
I think the key here is *not* to teach STEM people all the social science methodology and expect them to master it just as well as eg. programming. The key is to make them aware where their expertise ends, and value and *seek* the contributions of non-STEM experts.
The humanities & social sciences are not optional for STEM; they are integral. The longer institutions of higher education say that, but do the exact opposite in terms of funding & status for non-STEM fields, the more we’re going to screw up and the more we’re all going to suffer
1
7
53
I am hosting a roundtable this week at @ic4ai ! #AIethics #AIrisk #responsibleAI
2
51
If AI ethicists need a doctorate in CS to be qualified to critique AI, then AI researchers need to have a doctorate in ethics/humanities/philosophy/HCI/etc. to be qualified to build AI.
8
167
34
823
68,407
"There's no way a non-industry person can understand" On the contrary --- the folks in industry have shown themselves incapable of understanding (or caring about) the impacts of people of their tech. Regulation should protect rights and should be made by policymakers.
WATCH: Former Google CEO @ericschmidt tells #MTP Reports the companies developing AI should be the ones to establish industry guardrails — not policy makers. “There’s no way a non-industry person can understand what’s possible.”
14
147
2
544
69,042
No reasonable president would gather the CEOs of pharmaceutical companies and tell them "I hope you'll educate us as to what you think is most needed to protect society". They would first ask scientists who understand the risk-benefit balance and deal with CEOs for who they are.
Artificial Intelligence is one of the most powerful tools of our time, but to seize its opportunities, we must first mitigate its risks. Today, I dropped by a meeting with AI leaders to touch on the importance of innovating responsibly and protecting people's rights and safety.
14
82
4
353
57,633
It seems like we spend half our time talking to various legislators and agencies and STILL we have this shit. A room full of the dudes who gave us the issues & fired us for talking about the risks, being called on by the damn president to “protect people’s rights.”
Artificial Intelligence is one of the most powerful tools of our time, but to seize its opportunities, we must first mitigate its risks. Today, I dropped by a meeting with AI leaders to touch on the importance of innovating responsibly and protecting people's rights and safety.
34
227
12
868
145,800
This is a really good read about how generative AI imposes cultural norms present in the training data..but also why predicting emotions from facial expressions is pseudoscience #AIethics
I have been waiting for a reason to write about the American smile for decades. And AI finally gave it to me. AI and the American Smile How AI misrepresents culture through a facial expression. medium.com/@socialcreature/a…
1
2
118
Everyone’s talking about ChatGPT plugins, GPT-4 AGI sparks and I can only think about one thing: OpenAI has fooled us all. From non-profit open-source AI lab untethered from shareholders’ interests to the opposite in 8 years: closed-source, for-profit, fully corporate
117
711
118
4,103
1,015,403
It is very telling that when push comes to shove, despite having attracted some very talented, thoughtful, proactive, researchers, the tech cos decide they're better off without ethics/responsible AI teams. >>
4
27
135
11,156
Twitter to me on my first tweet of the day 🤦‍♀️
2
72
"The compulsion to show 'balance' by always referring to AI's alleged potential for good should be dropped by acknowledging that the social benefits are still speculative while the harms have been empirically demonstrated." -- @danmcquillan >>
2
20
1
85
8,989
Oh boy am I mad... and will definitely tell this story whenever I hear this term in the future! TL;DR it was made up by a Psychiatrist to discredit a woman (who was held hostage by a bankrobber) who questioned his authority 🙄
I just read -- in the book SEE WHAT YOU MADE ME DO by Jess Hill -- about the incident in Stockholm for which Stockholm Syndrome is named, and I am royally pissed off on behalf of all women, buckle up.
75
"Once people learned the messages were co-created by a machine, it didn’t work. Simulated empathy feels weird, empty." 🤦‍♀️Not only that, but it's extremely dangerous to pretend a language-parroting algorithm could provide genuine support in a mental health crisis!
We provided mental health support to about 4,000 people — using GPT-3. Here’s what happened 👇
2
2
188
They pulled the feature... but really, was this a necessary experiment?
1
30
In which Santa's elves learn an important less about AI bias. 🎅🤖🧵 'Twas the weeks before Christmas And Santa was busy His list was so long It was making him dizzy There are too many children List checking can’t scale If I do this by hand I will certainly fail! 1/n
2
16
7
55
21,464
Surprise surprise, the newest internet toy is not free of bias, and doesn't even have solid mitigation in place.
Yes, ChatGPT is amazing and impressive. No, @OpenAI has not come close to addressing the problem of bias. Filters appear to be bypassed with simple tricks, and superficially masked. And what is lurking inside is egregious. @Abebab @sama tw racism, sexism.
2
3
I asked ChatGPT to rewrite Bohemian Rhapsody to be about the life of a postdoc, and the output was flawless:
133
1,577
382
8,698
Or we could reroute funding toward Social Sciences instead, and make sure to collaborate with researchers who already are experts with years of training and research on these topics?
Most computer scientists (myself included) have shallow social sciences knowledge. I argue that #AI researchers need significant social science training to understand and empathize with complex issues, like bias and equity, and may be even history and law.
5
7
Afraid of getting pregnant? Maybe consider changing your name! 🤦‍♀️🤦‍♀️ “With technology, based on name, surname and address, you can predict five or six years ahead which girl, or future teenager, is 86% predestined to have a teenage pregnancy” giswatch.org/sites/default/f…
"The idea that algorithms can predict teenage pregnancy before it happens is the perfect excuse for anti-women and anti-sexual and reproductive rights activists to declare abortion laws unnecessary. According to their narratives, if they have enough information from poor families
2
1
2
Cannot believe this has to be said: if it is **impossible** to do, then no algorithm in this world can do it either. Especially not those named "AI". No, not even if they have a picture of your face...-.-
2