Topics
Latest
AI
Amazon
Image Credits:TechCrunch
Apps
Biotech & Health
mood
Image Credits:TechCrunch
Cloud Computing
Department of Commerce
Crypto
Enterprise
EVs
Fintech
fundraise
Gadgets
Gaming
Government & Policy
computer hardware
Layoffs
Media & Entertainment
Meta
Microsoft
concealment
Robotics
security system
Social
Space
startup
TikTok
expatriation
Venture
More from TechCrunch
Events
Startup Battlefield
StrictlyVC
Podcasts
Videos
Partner Content
TechCrunch Brand Studio
Crunchboard
touch Us
To giveAI - focused womenacademics and others their well - deserved — and overdue — time in the spotlight , TechCrunch is launching aseries of interviewsfocusing on noteworthy women who ’ve contributed to the AI revolution . We ’ll release several slice throughout the year as the AI boom continues , play up key work that often goes unrecognized . Read more profileshere .
Heidy Khlaaf is an technology manager at the cybersecurity firm Trail of Bits . She specializes in evaluate software and AI implementations within “ base hit vital ” system of rules , like nuclear power plants and sovereign vehicles .
Khlaaf pick up her computing machine science PhD from the University College London and her b in electronic computer science and philosophy from Florida State University . She ’s led guard and security audited account , provided consultations and reviews of sureness cases and contributed to the conception of standard and guidelines for safety- and certificate - related applications and their growth .
Q&A
in brief , how did you get your start in AI ? What attracted you to the field ?
I was drawn to robotics at a very young age and started programming at the age of 15 , as I was fascinated with the prospect of using robotics and AI ( as they ’re inexplicably link ) to automate workloads where they ’re most needed . Like in manufacturing , I saw robotics being used to help the elderly — and automate life-threatening manual labor in our society . I did , however , receive my PhD in a unlike subfield of computer scientific discipline , because I think that having a unattackable theoretical grounding in computer science permit you to make educated and scientific conclusion into where AI may or may not be desirable , and where pitfall may be .
What employment are you most lofty of in the AI field ?
Using my strong expertness and background in safety engine room and safety - critical organization to put up context and literary criticism where needed on the new field of AI “ safe . ” Although the discipline of AI safety has essay to adjust and advert well - established guard and security proficiency , various terminology has been misconstrued in its role and import . There is a lack of reproducible or knowing definition that do compromise the integrity of the safety techniques the AI community is currently using . I ’m particularly proud of “ Toward Comprehensive Risk Assessments and Assurance of AI - Based Systems ” and “ A Hazard Analysis Framework for Code Synthesis Large Language Models ” where I deconstruct false narration about safety and AI evaluations and provide concrete steps on bridging the prophylactic opening within AI .
Join us at TechCrunch Sessions: AI
Exhibit at TechCrunch Sessions: AI
How do you pilot the challenge of the male person - dominated tech industriousness and , by extension , the male person - overshadow AI manufacture ?
Acknowledgment of how little the condition quo has changed is not something we discuss often , but I believe is actually important for myself and other technological women to understand our post within the diligence and hold a naturalistic view on the changes required . Retention rates and the ratio of women hold in leadership lieu has remained largely the same since I joined the domain , and that ’s over a decade ago . And as TechCrunch has capably orient out , despite awful breakthroughs and contribution by women within AI , we continue sidelined from conversations that we ourselves have defined . recognise this deficiency of progress helped me understand that build a solid personal biotic community is much more worthful as a seed of support rather than bank on DEI initiatives that unfortunately have not moved the needle , given that prejudice and skepticism toward technical women are still quite permeative in technical school .
What advice would you give to womanhood seeking to enter the AI field ?
Not to appeal to authority and to regain a line of workplace that you genuinely believe in , even if it contravene pop story . give the power AI labs hold politically and economically at the moment , there is an inherent aptitude to take anything AI “ thought leaders ” United States Department of State as fact , when it is often the case that many AI claims are marketing speak that overstate the abilities of AI to profit a bottom line of merchandise . Yet , I see pregnant hesitancy , especially among junior fair sex in the field of study , to vocalize skepticism against claims made by their male peers that can not be substantiated . Imposter syndrome has a inviolable detention on charwoman within tech and leads many to doubt their own scientific integrity . But it is more important than ever to challenge claim that exaggerate the capability of AI , peculiarly those that are not confirmable under the scientific method acting .
What are some of the most pressing result facing AI as it develop ?
no matter of the advancements we ’ll honor in AI , they will never be the singular solution , technologically or socially , to our issues . Currently there is a vogue to shoehorn AI into every potential system , disregarding of its potency ( or lack thence ) across legion domains . AI should augment human capabilities rather than replace them , and we are witnessing a complete neglect of AI ’s pitfalls and failure modes that are lead to real real hurt . Just late , an AI systemShotSpotterrecently head to an officer firing at a child .
What are some issues AI users should be aware of ?
How truly unreliable AI is . AI algorithms are notoriously flawed with high mistake rates observed across applications that require preciseness , accuracy and refuge - criticality . The direction AI systems are trained embed human bias and favoritism within their outputs that become “ de facto ” and automatise . And this is because the nature of AI scheme is to provide event establish on statistical and probabilistic inference and correlations from diachronic data , and not any type of abstract thought , actual evidence or “ causing . ”
What is the good agency to responsibly construct AI ?
To ensure that AI is develop in a way that protects citizenry ’s right and safety through constructing verifiable claim and bear AI developers accountable to them . These claims should also be scoped to a regulatory , condom , honourable or technological diligence and must not be verifiable . Otherwise , there is a significant lack of scientific integrity to fittingly value these system . Independent regulators should also be measure AI systems against these claim as currently required for many mathematical product and system in other industries — for model , those judge by the FDA . AI system should not be nontaxable from stock auditing mental process that are well - established to control public and consumer protection .
How can investor considerably advertise for responsible AI ?
investor should engage with and investment company organizations that are seeking to establish and advance auditing drill for AI . Most funding is currently indue in AI research lab themselves , with the belief that their condom teams are sufficient for the advancement of AI valuation . However , independent auditor and regulators are cardinal to public trust . Independence allows the public to trust in the accuracy and wholeness of assessments and the integrity of regulatory outcomes .
