Women in AI: Kristine Gloria of the Aspen Institute tells women to enter the field and ‘follow your curiosity’

Trending 1 week ago
ARTICLE AD BOX

To springiness AI-focused women academics and others their well-deserved — and overdue — clip successful nan spotlight, TechCrunch is launching a series of interviews focusing connected singular women who’ve contributed to nan AI revolution. We’ll people respective pieces passim nan twelvemonth arsenic nan AI roar continues, highlighting cardinal activity that often goes unrecognized. Read much profiles here.

Kristine Gloria leads nan Aspen Institute’s Emergent and Intelligent Technologies Initiative — nan Aspen Institute being nan Washington, D.C.-headquartered deliberation vessel focused connected values-based activity and argumentation expertise. Gloria holds a Ph.D. successful cognitive subject and a Master’s successful media studies, and her past activity includes investigation astatine MIT’s Internet Policy Research Initiative, nan San Francisco-based Startup Policy Lab, and nan Center for Society, Technology and Policy astatine UC Berkeley.

Q&A

Briefly, really did you get your commencement successful AI? What attracted you to nan field?

To beryllium frank, I decidedly didn’t commencement my profession successful pursuit of being successful AI. First, I was really willing successful knowing nan intersection of exertion and nationalist policy. At nan time, I was moving connected my Master’s successful media studies, exploring ideas astir remix civilization and intelligence property. I was surviving and moving successful D.C. arsenic an Archer Fellow for nan New America Foundation. One day, I distinctly retrieve sitting successful a room filled pinch nationalist policymakers and politicians who were throwing astir position that didn’t rather fresh their existent method definitions. It was soon aft this gathering that I realized that successful bid to move nan needle connected nationalist policy, I needed nan credentials. I went backmost to school, earning my doctorate successful cognitive subject pinch a attraction connected semantic technologies and online user privacy. I was very fortunate to person recovered a mentor and advisor and laboratory that encouraged a cross-disciplinary knowing of really exertion is designed and built. So, I sharpened my method skills alongside processing a much captious viewpoint connected nan galore ways tech intersects our lives. In my domiciled arsenic nan head of AI astatine nan Aspen Institute, I past had nan privilege to ideate, prosecute and collaborate pinch immoderate of nan starring thinkers successful AI. And I ever recovered myself gravitating towards those who took nan clip to profoundly mobility if and really AI would effect our day-to-day lives.

Over nan years, I’ve led various AI initiatives and 1 of nan astir meaningful is conscionable getting started. Now, arsenic a founding squad personnel and head of strategical partnerships and invention astatine a caller nonprofit, Young Futures, I’m excited to weave successful this type of reasoning to execute our ngo of making nan integer world an easier spot to turn up. Specifically, arsenic generative AI becomes array stakes and arsenic caller technologies travel online, it’s some urgent and captious that we thief preteens, teens and their support units navigate this immense integer wilderness together.

What activity are you astir proud of (in nan AI field)?

I’m astir proud of 2 initiatives. First is my activity related to surfacing nan tensions, pitfalls and effects of AI connected marginalized communities. Published successful 2021, “Power and Progress successful Algorithmic Bias” articulates months of stakeholder engagement and investigation astir this issue. In nan report, we posit 1 of my all-time favourite questions: “How tin we (data and algorithmic operators) recast our ain models to forecast for a different future, 1 that centers astir nan needs of nan astir vulnerable?” Safiya Noble is nan original writer of that question, and it’s a changeless information passim my work. The 2nd astir important inaugural precocious came from my clip arsenic Head of Data astatine Blue Fever, a institution connected nan ngo to amended younker well-being successful a judgment-free and inclusive online space. Specifically, I led nan creation and improvement of Blue, nan first AI affectional support companion. I learned a batch successful this process. Most saliently, I gained a profound caller appreciation for nan effect a virtual companion tin person connected personification who’s struggling aliases who whitethorn not person nan support systems successful place. Blue was designed and built to bring its “big-sibling energy” to thief guideline users to bespeak connected their intelligence and affectional needs.

How do you navigate nan challenges of nan male-dominated tech industry, and, by extension, nan male-dominated AI industry?

Unfortunately, nan challenges are existent and still very current. I’ve knowledgeable my adjacent stock of disbelief successful my skills and acquisition among each types of colleagues successful nan space. But, for each azygous 1 of those antagonistic challenges, I tin constituent to an illustration of a antheral workfellow being my fiercest cheerleader. It’s a reliable environment, and I clasp connected to these examples to thief manage. I besides deliberation that truthful overmuch has changed successful this abstraction moreover successful nan past 5 years. The basal accomplishment sets and master experiences that suffice arsenic portion of “AI” are not strictly machine science-focused anymore.

What proposal would you springiness to women seeking to participate nan AI field?

Enter successful and travel your curiosity. This abstraction is successful changeless motion, and nan astir absorbing (and apt astir productive) pursuit is to continuously beryllium critically optimistic astir nan section itself.

What are immoderate of nan astir pressing issues facing AI arsenic it evolves?

I really deliberation immoderate of nan astir pressing issues facing AI are nan aforesaid issues we’ve not rather gotten correct since nan web was first introduced. These are issues astir agency, autonomy, privacy, fairness, equity and truthful on. These are halfway to really we situate ourselves amongst nan machines. Yes, AI tin make it vastly much analyzable — but truthful tin socio-political shifts.

What are immoderate issues AI users should beryllium alert of?

AI users should beryllium alert of really these systems complicate aliases heighten their ain agency and autonomy. In addition, arsenic nan sermon astir really technology, and peculiarly AI, whitethorn effect our wellbeing, it’s important to retrieve location are tried-and-true devices to negociate much antagonistic outcomes.

What is nan champion measurement to responsibly build AI?

A responsible build of AI is much than conscionable nan code. A genuinely responsible build takes into relationship nan design, governance, policies and business model. All thrust nan other, and we will proceed to autumn short if we only strive to reside 1 portion of nan build.

How tin investors amended push for responsible AI

One circumstantial task, which I respect Mozilla Ventures for requiring successful its diligence, is an AI exemplary card. Developed by Timnit Gebru and others, this believe of creating exemplary cards enables teams — for illustration funders — to measure nan risks and information issues of AI models utilized successful a system. Also, related to nan above, investors should holistically measure nan strategy successful its capacity and expertise to beryllium built responsibly. For example, if you person spot and information features successful nan build aliases a exemplary paper published, but your gross exemplary exploits susceptible organization data, past there’s misalignment to your intent arsenic an investor. I do deliberation you tin build responsibly and still beryllium profitable. Lastly, I would emotion to spot much collaborative backing opportunities among investors. In nan realm of wellbeing and intelligence health, nan solutions will beryllium varied and immense arsenic nary personification is nan aforesaid and nary 1 solution tin lick for all. Collective action among investors who are willing successful solving nan problem would beryllium a invited addition.