Topics
late
AI
Amazon
Image Credits:Bryce Durbin/TechCrunch
Apps
Biotech & Health
Climate
Image Credits:Bryce Durbin/TechCrunch
Cloud Computing
Commerce
Crypto
enterprisingness
EVs
Fintech
Fundraising
gizmo
Gaming
Government & Policy
Hardware
Layoffs
Media & Entertainment
Meta
Microsoft
Privacy
Robotics
certificate
Social
distance
startup
TikTok
Transportation
Venture
More from TechCrunch
result
Startup Battlefield
StrictlyVC
Podcasts
video
Partner Content
TechCrunch Brand Studio
Crunchboard
Contact Us
To giveAI - centre womenacademics and others their well - deserve — and delinquent — time in the spotlight , TechCrunch is launching aseries of interviewsfocusing on remarkable women who ’ve contributed to the AI rotation . We ’ll publish part throughout the year as the AI gold rush continues , highlighting key work that often goes unrecognized . Read more profileshere .
Emilia Gómez is a principal investigator at the European Commission ’s Joint Research Centre and scientific coordinator of AI Watch , the EC first step to monitor the advancements , intake and impingement of AI in Europe . Her team contributes with scientific and technical cognition to EC AI policies , include the recently proposedAI Act .
Gómez ’s research is run aground in the computational music field , where she contributes to the discernment of the way humans describe music and the method acting in which it ’s modeled digitally . Starting from the euphony field , Gómez investigates the impingement of AI on human behavior — in particular the effects on jobs , decisions and tyke cognitive and socioemotional development .
Q&A
Briefly , how did you get your scratch in AI ? What attracted you to the field of operations ?
I started my enquiry in AI , in especial in automobile learning , as a developer of algorithms for the automatic description of music audio recording sign in terms of melody , tonality , law of similarity , style or emotion , which are tap in different applications from music platforms to pedagogy . I started to search how to design novel machine see approach shot make do with different computational chore in the music field , and on the relevancy of the data pipeline include datum go under world and note . What I liked at the time from machine learning was its modelling capableness and the shift from knowledge - driven to data - driven algorithm design — for example or else of designing descriptor based on our knowledge of acoustics and medicine , we were now using our know - how to design data sets , architecture and breeding and evaluation procedure .
From my experience as a machine learning research worker , and find my algorithmic program “ in military action ” in different field , from music platform to symphonic music concerts , I realized the huge impact that those algorithmic rule have on people ( for instance listeners , musician ) and directed my research toward AI evaluation rather than development , in particular on studying the wallop of AI on human behavior and how to evaluate system in terms of aspects such as equity , human supervision or foil . This is my squad ’s current enquiry topic at the Joint Research Centre .
What work are you most gallant of ( in the AI battleground ) ?
Join us at TechCrunch Sessions: AI
Exhibit at TechCrunch Sessions: AI
On the academic and proficient side , I ’m proud of my contribution to music - specific machine learning architecture at the Music Technology Group in Barcelona , which have advanced the state of the artistic creation in the force field , as it ’s reflected in my citation records . For illustration , during my PhD I propose a data - drive algorithm to educe tonality from audio signals ( for example if a melodious man is in C major or five hundred minor ) which has become a key reference in the battlefield , and later I co - designed machine learning methods for the automatic description of music sign in terminus of melody ( e.g. used to search for songs by humming ) , tempo or for the modeling of emotions in music . Most of these algorithms are presently integrated into Essentia , an subject source library for audio and euphony analysis , verbal description and synthesis and have been exploited in many recommender organization .
I ’m especially proud of Banda Sonora Vital ( LifeSoundTrack ) , a project grant by Red Cross Award for Humanitarian Technologies , where we developed a individualised music recommender adapted to aged Alzheimer patients . There ’s also PHENICX , a large European Union ( EU)-funded project I ordinate on the consumption of music ; and AI to create enriched symphonious euphony experiences .
I love the music computing residential area and I was happy to become the first distaff President of the United States of the International Society for Music Information Retrieval , to which I ’ve been give all my vocation , with a particular interest in increasing multifariousness in the battleground .
presently , in my purpose at the Commission , which I joined in 2018 as lead scientist , I provide scientific and proficient support to AI policies arise in the EU , notably the AI Act . From this recent work , which is less visible in condition of publications , I ’m proud of my lowly proficient part to the AI Act — I say “ humble ” as you may guess there are many people involved here ! As an example , there ’s a tidy sum of work I contributed to on the harmonization or translation between legal and technological terms ( e.g. proposing definitions grounded in subsist literature ) and on assessing the practical effectuation of legal requirement , such as transparency or technical certification for gamy - hazard AI systems , general - aim AI models and generative AI .
I ’m also quite proud of my team ’s body of work in stomach the EU AI indebtedness directive , where we studied , among others , particular characteristic that make AI systems inherently speculative , such as deficiency of causality , opacity , capriciousness or their self- and continuous - learning capacity , and assess associated difficulties show when it arrive to proving causation .
How do you sail the challenge of the male person - dominated tech diligence , and , by extension , the male - dominate AI diligence ?
It ’s not only technical school — I ’m also navigating a male - overshadow AI research and policy field ! I do n’t have a technique or a strategy , as it ’s the only environment I know . I do n’t know how it would be to work in a diverse or a female - overlook workings surround . “ Would n’t it be overnice ? , ” like the Beach Boys ’ song goes . I honestly render to avoid thwarting and have fun in this challenging scenario , working in a world prevail by very assertive guys and enjoying collaborating with fantabulous women in the field of force .
What advice would you give to womanhood seeking to enter the AI domain ?
I would tell them two thing :
You ’re much needed — please enter our field , as there ’s an urgent need for diversity of visions , approach and ideas . For illustration , fit in to the divinAI labor — a project I co - founded on monitoring diversity in the AI airfield — only 23 % of author epithet at the International Conference on Machine Learning and 29 % at the International Joint Conference on AI in 2023 were female , regardless of their gender identity .
You are n’t alone — there are many women , nonbinary colleagues and manlike allies in the field , even though we may not be so visible or recognise . see for them and get their mentoring and keep ! In this circumstance , there are many phylogenetic relation groups present in the research theatre . For instance , when I became prexy of the International Society for Music Information Retrieval , I was very active in the Women in Music Information Retrieval go-ahead , a pioneer in diversity efforts in music computing with a very successful mentoring program .
What are some of the most urgent issues confront AI as it evolve ?
In my opinion , researchers should consecrate as many efforts to AI development as to AI evaluation , as there ’s now a want of residual . The research community is so engaged advance the state of the art in full term of AI capableness and performance and so excited to see their algorithm used in the veridical world that they draw a blank to do right evaluations , shock assessment and external audit . The more levelheaded AI organization are , the more intelligent their evaluations should be . The AI evaluation landing field is under - canvass , and this is the cause of many incidents that give AI a bad reputation , e.g. sex or racial bias present in datum sets or algorithmic rule .
What are some issues AI drug user should be aware of ?
citizen using AI - powered tools , like chatbots , should know that AI is not magic . Artificial intelligence is a Cartesian product of human intelligence activity . They should teach about the wreak principles and limitations of AI algorithms to be able to challenge them and utilize them in a responsible manner . It ’s also important for citizen to be inform about the quality of AI product , how they are appraise or certified , so that they know which ones they can desire .
What is the best way to responsibly make AI ?
In my view , the dependable way to develop AI products ( with a just societal and environmental impingement and in a responsible way ) is to pass the needed resource on rating , assessment of social impact and mitigation of risks — for instance , to rudimentary rights — before place an AI system in the grocery store . This is for the benefit of business and reliance on products , but also of lodge .
Responsible AI or trusty AI is a way to ramp up algorithms where aspects such as transparency , fairness , human superintendence or social and environmental well - being indigence to be addressed from the very start of the AI design summons . In this sense , the AI Act not only congeal the bar for regulating hokey news worldwide , but it also reflects the European emphasis on trustworthiness and transparence — enabling innovation while protecting citizens ’ rights . This I palpate will increase citizen cartel in the product and technology .
