Founder & CEO @MindfulAILab | AI Ethics Advisory & Audits | Responsible AI Strategy | Past Fellow @BKCHarvard @mozilla @datasociety | #100BrilliantWomenInAI
- Tweets 180
- Following 1,421
- Followers 1,651
- Likes 885
Pinned Tweet
Honored to be part of this group of amazing group of women doing critically important work!
(10/10) The moment you all have been waiting for… Unveiling the #100BrilliantWomen in AI Ethics™️ list for 2021! @sherrying @suchanaseth @susanleavy @kite_kite_ @TanDuarte @Combsthepoet @tinaeliassi @toni_erskine @tulsiparida 100brilliantwomeninaiethics.…
ALT 100 Brilliant Women in AI Ethics™ 2021 List: Serife Wong, Suchana Seth, Susan Leavy, Suzanne Kite, Tania Duarte, Tawana Petty, Tiffany Deng, Tina Eliassi-Rad, Toni Erskine, Tulsi Parida
67
13
39
Let's talk about #auditability and #accountability in the context of India's draft Non-Personal Data framework today at 6 pm with @hasgeek
Thanks @zainabbawa for making this important conversation happen.
hasgeek.com/PrivacyMode/npd-…
17
8
20
I'm co-hosting a conversation tonight on putting FAccT into machine learning practice with @scribbledata supported by @fifthel at @hasgeek
hasgeek.com/fifthelephant/op…
Thanks @zainabbawa @pingali @Guava_man for putting this together.
4
8
New episode on Listen! @suchanaseth on different definitions of fairness in the context of machine learning. buff.ly/2rVHkbL
1
4
8
My favorite sign at #marchforscience @teon_io @daniellecrobins @kirstie_j @bmpvieira @auremoser @shefw
I probably won't be able to make it to the @march4sciencela. But I made a poster for those who can!!!! #GrabEmByTheData #ScienceMarch
2
1
8
Read. How attention economy business models shape the ethics of tech platforms. washingtonmonthly.com/magazi…
1
7
Thanks @alokpi for inviting me to write this. epw.in/journal/2017/51/priva… nitter.app/alokpi/status/94453623…
3
7
"We expect simple interpretable ML models to be monotonic, even if they are non-linear." @jpatrickhall talking at #OReillyAI #ResponsibleML
3
1
5
Tuned in to the #antitrusthearing ... Left because it sounds like an illustration of #AnIllustratedBookOfBadArguments by @alialmossawi ... Catching up on my @rightscon tech demo sessions instead.
1
4
We need positive examples of businesses monetizing data ethically. Thanks @DarkSkyApp for your perspective.
Wow. This is why we need to find business solutions to safeguard privacy: blog.darksky.net/location-pr… (Thanks @suchanaseth for the link)
2
5
Summary stats on steroids. I have wanted this R package forever! skimr's output is pipable, and computable. Much love!
[blog] `skimr` #rstats 📦 for useful and tidy summary statistics ropensci.org/blog/blog/2017/… by @juliesquid #runconf17 1/2
4
4
Really clever approach to finding near-optimal solutions for the asymmetric traveling salesman problem. quantamagazine.org/one-way-s…
2
4
Deep learning to predict ground state energy of an electron under potentials without analytic form!
journals.aps.org/pra/abstrac…
1
4
Incredibly excited to build @MindfulAILab - come talk to us about all things #AlgorithmAudit and #AIEthics
The Mindful AI Lab offers AI ethics advisory services, audits AI systems and data practices, and empowers humans to build ethical AI. Check out mindfulailab.com for more.
#TheMindfulAILab #AI #ArtificalIntelligence #ResponsibleAI #AIEthics #AIAudit
3
Do we want to build a world where we can charge our phone by yelling at it, even if we can? @mcette talking at #OReillyAI @datasociety
4
“We leave open how to interpret the human subjects’ responses that ranked CAN art better than the Art Basel samples" technologyreview.com/s/60819…
1
1
3
If you're going to read one paper about #ResponsibleAI or #MachineLearning best practices this year, then this paper by Bin Yu is THE one! pnas.org/content/117/8/3893
1
3
3
Increase font size in axis labels. Use marginalia to write xkcd style comparisons for each order of magnitude jump on axis.
1
3
Wonderful discussion around model stacking and "emergent ethical problems" in IoT at the @VIRT_EU advisor meeting. @quillis @a_b_powell
4
2
Watch me talk about putting AI ethics into practice - doing data science responsibly.
hasgeek.com/fifthelephant/ma…
Thanks @zainabbawa and @scribbledata for making this happen.
1
2
3
"Can we view fairness as a system level property of ML systems rather than a component reliability issue?" #OReillyAI #ResponsibleML
At 4:00 at #OReillyAI, #MITREcorp's Lashon Booker, Chuck Howell present on accelerants, enablers of AI adoption. oreil.ly/2sMBsQH
1
2
3
#Aadhaar might need our blood/DNA samples next? Unless we challenge that in court? Who made @UIDAI the owner of our bodies?!?
J. Chandrachud: "Biological attributes" is open ended.
AG says that blood, urine, dna can be added, but it'll be subject to examination by the courts, just like right now the court is examining whether collection of fingerprints and Iris scans are a violation of privacy.
4
2
@geminiimatt @blackswanburst copyright felonizes reverse engineering?!? @VIRT_EU design considerations for ethical IoT?
John Deere just told the copyright office that only corporations can own property, humans can only license it
boingboing.net/2017/04/22/dr…
2
3
3
This is getting to be my favorite "prank the neural net" thread ever! nitter.app/jahnnie/status/9697357…
1
5
1
Great data sonification + awesome new work music @tenacioustek @VenkatramananPR Thanks @daniellecrobins
2
"Lots of people want to work on research distillation. Unfortunately, it's very difficult to do so, because we don't support them." So true!
Announcing Distill (@distillpub), a science journal and ecosystem supporting the understanding of #MachineLearning - goo.gl/yB5Yl0
1
1
"... although limiting location access sometimes breaks features within apps." Treating #privacy as commodity rather than right. #DataEthics
The Internet’s most dominant business model of selling your - here location - data for ‘free’ or ‘cheap’ has moved to the analogue theater world. With that will follow more distrust 9to5mac.com/2018/03/05/movie…
5
2
Watch Tishby's talk on information bottleneck and deep learning.
piped.video/watch?v=bLqJHjXi…
1
2
Thank you @block76
aimyths.org/ is an excellent resource put together by @mozilla fellow @djleufer
Are there better measures of user utility than normalized discounted cumulative gain for fair recommendation systems? @mdekstrand #FAT2018
2
Thanks @kirstie_j for being an amazing host, and the audience @turinginst for some great questions!
1
2
This is awesome!
Yes! - we drafted up something for CS 1 courses this summer. medium.com/bucknell-hci/the-… Are there other examples?
2
We absolutely need to talk about translating values and ethical principles into practice! And the practical challenges to doing it! And how to teach data science ethics!
2
2
Love this by @zararah Ever struggled to answer the dreaded "So where are you from?" Read this. reallifemag.com/where-are-yo…
1
2
"You fix what you find, and you find what you look for." #ResponsibleML #OReillyAI
1
2
Cool new snitching portal from the makers of Taser arstechnica.com/tech-policy/…
1
@JennyBryan on working open, juggling motherhood and work, and the vital but "unglamorous" work of building and teaching good workflow tooling for data science
1
The gap between "open source" and "transparent". nitter.app/staeiou/status/9210204…
1
1
So heartening to see the enthusiasm around making machine learning algorithms fairer. #RightsCon #algorithms #ethicalml #aiethics
1
1
"Train an interpretable model - like a short decision tree - to mimic or surrogate the behaviour of your complex model."
1
1
Looking forward to reading chapter 13 - especially connections with operator theory and group theory.
Avi Wigderson's new book draft "Mathematics and Computation" now available online math.ias.edu/avi/book
A great tour of complexity.
1
Next on my reading list #scalefreenetworks #statisticalphysics
Thanks, @PhysicsToday for chatting with me about The Diversity of Networks: ow.ly/Lu2e30bac6Q
1
Murderbot series by Martha Wells for a good laugh
Ambiguity Machines by Vandana Singh
Ancilliary Justice by Ann Leckie
Cox by Christoph Ransmayr for a more alt history / speculative fiction feel
1
1
"You want to know what the machine knows..." @KJ_Hammond speaking on AI for natural language generation at #OReillyAI
1
Loving Julia Reda's Copywrongs 2.0 talk at #33c3 training ML models may become illegal?!? @VenkatramananPR @datasociety @gilgul
1
1