Yesterday, my giant looming head joined The Royal Society launch of their Science in the Age of AI report, which is full of brilliant recommendations and insights. In Areeq Chowdhury's words: It warns that an overdependence on 'opaque' AI systems could make scientific research findings less reliable & limit their usefulness for solving real world challenges. This does not stop AI generating useful insights, but there are various significant challenges to overcome. Some more favorite pieces: - public infrastructure in compute and usable data is paramount - better access mechanisms for high quality and sensitive data - research methodologies as an artifact for release - investing in domain-specific taxonomies - labo(u)r as infrastructure - better understanding tradeoffs of "performance" and "safety" - reproducibility varies across cultural contexts Congrats to the team and thanks to Areeq Chowdhury, Alison Noble, Dawn Bloxwich, and Max Beverton-Palmer for a great discussion! Read the report here: https://lnkd.in/gTnrMam4
Irene Solaiman’s Post
More Relevant Posts
-
President/CEO at SanTrac Technologies/Managed IT-Cloud IT-CyberSecurity-VoIP-Low Voltage Communications Provider/ Author of "The Secret To Finding Honest Competent Responsive IT Services"
🌐 AI shows promise, but AI bias remains a challenge, reflecting societal prejudices. 🌟 🔧 Key steps towards equality in AI: diversify perspectives, engage interdisciplinary experts, and commit to education. 👉 Read the full blog here! https://conta.cc/3XhfVON
To view or add a comment, sign in
-
Exciting times in AI! Let's celebrate the remarkable work of Francine Bennett, who champions making AI responsible and inclusive. From finding medical treatments with AI in biotech to her leadership at the Ada Lovelace Institute, Francine exemplifies the impact of putting people and society at the heart of technology. For those passionate about AI's future, her advice is golden: "Enjoy it! ...start by starting on something you’re intrigued by, and work from there." As AI evolves, ensuring it works for everyone remains crucial. Let's engage in building AI that's responsible, ethical, and truly beneficial for all. Share your thoughts and experiences on how we can shape an inclusive AI future together. #AI #WomenInTech #ResponsibleAI #InclusiveTechnology #UKTech
To view or add a comment, sign in
-
The importance of diversity in the field of artificial intelligence (AI) cannot be overstated. A more diverse workforce brings a wider range of perspectives, ideas and experiences to the table, which can lead to more innovative and inclusive solutions. This is particularly important in the development and deployment of AI technologies, which have the potential to significantly impact public services and the communities they serve. A lack of diversity in AI can lead to algorithms and systems that perpetuate existing biases and inequalities, rather than addressing them. This can have serious consequences for marginalized communities, who may be disproportionately affected by biased AI systems. To ensure that AI technologies are developed and used in a way that benefits all members of society, it is essential to prioritize diversity and inclusion in the field. This includes increasing representation of underrepresented groups in AI research and development, as well as ensuring that the perspectives and needs of diverse communities are taken into account in the design and implementation of AI systems. By fostering a more diverse and inclusive AI community, we can harness the full potential of this powerful technology to improve public services and create a more equitable society for all. read about our services @ waiu.org #AI #Diversity #Inclusion #PublicServices #Equity #digitaltransformation #oman Loai A.
Home | World AI University
waiu.org
To view or add a comment, sign in
-
Must-watch interview with 𝗚𝗲𝗼𝗳𝗳𝗿𝗲𝘆 𝗛𝗶𝗻𝘁𝗼𝗻, one of the godfathers of AI. While most people, including him, think the overall outcome can be positive, he emphasizes the 𝗻𝗲𝗰𝗲𝘀𝘀𝗮𝗿𝘆 𝗽𝗿𝗲𝗰𝗮𝘂𝘁𝗶𝗼𝗻𝘀 𝗳𝗼𝗿 𝗔𝗜. 𝗞𝗲𝘆 𝗧𝗮𝗸𝗲𝗮𝘄𝗮𝘆𝘀: 1. 𝗔𝗜 𝗶𝗺𝗽𝗿𝗼𝘃𝗲𝗺𝗲𝗻𝘁 is driven by larger computational capacities and scientific breakthroughs (e.g., transformers 2017). 2. AI can reach 𝘀𝘂𝗽𝗲𝗿𝗶𝗻𝘁𝗲𝗹𝗹𝗶𝗴𝗲𝗻𝗰𝗲 within the next 20-30 years. 🧠 3. AI will evolve with "Self-preservation" ➡ "Self-interest" ➡ "𝗘𝘃𝗼𝗹𝘂𝘁𝗶𝗼𝗻" ➡ Competition will determine the winner. 🏆 4. Intelligent entities are rarely controlled by less intelligent ones, posing possible 𝘀𝗶𝗴𝗻𝗶𝗳𝗶𝗰𝗮𝗻𝘁 𝗿𝗶𝘀𝗸𝘀 to humans. ⚠️ 5. In the race for profits and power, AI companies may neglect safety. 💸 6. AI companies should allocate substantial resources (e.g., 33%) to 𝘀𝗮𝗳𝗲𝘁𝘆 𝗺𝗲𝗮𝘀𝘂𝗿𝗲𝘀. 🛡️ 7. Hinton believes there is a significant chance (even up to 50%) that AI developments could lead to significantly negative outcomes. 🔮 In my opinion, we are witnessing a transformative event that 𝘀𝘂𝗿𝗽𝗮𝘀𝘀𝗲𝘀 𝗮𝗻𝘆 𝗽𝗿𝗲𝘃𝗶𝗼𝘂𝘀 𝘁𝗲𝗰𝗵𝗻𝗼𝗹𝗼𝗴𝗶𝗰𝗮𝗹 𝗮𝗱𝘃𝗮𝗻𝗰𝗲𝗺𝗲𝗻𝘁 in human history. However, it is mistakenly being compared to the invention of computers or the Internet, yet few are discussing its 𝘁𝗿𝘂𝗲 𝗳𝘂𝘁𝘂𝗿𝗲 𝗶𝗺𝗽𝗮𝗰𝘁 𝗼𝗻 𝗵𝘂𝗺𝗮𝗻𝗶𝘁𝘆. 🌍 Thanks, Jon Erlichman, for the great interview!
To view or add a comment, sign in
-
🧮 Walk the Talk with AI 🧮 Grateful I was invited to join a discussion on AI's real-world applications in diverse industries. What I learned: - a machine learning algorithm that is part of an FDA approved application can’t be just retrained when new data is available but needs to go through the whole approval process again. - it would actually be more interesting if we named concrete AI methods and approaches instead of the generic buzzword „artificial intelligence“. - it is not always clear how much of an application‘s value is actually generated by AI technologies vs. traditional data analysis or even normal software. Thank you Katka Letzing, Kickstart Innovation and AI House Davos. Rudolf M. Moos, Hirslanden John Klepper, PIPRA Markus Wolf, Die Mobiliar Tina Willibald, Swisscom Ventures Benjamin Huber-Steinemann, viboo João Pedro (JP) Monteiro, Veezoo
To view or add a comment, sign in
-
Keynote Speaker & AI Strategy Consultant | Empowering Organizations to Navigate the Complexities of AI Implementation
Will.I.Am's AI Revolution: Empowering Diversity at CogX Festival in LA Last week, I had the privilege of attending the inaugural CogX Festival (usually based in the UK) in Los Angeles, where I witnessed many discussions about the future of technology. The event highlighted how AI will impact the future of work, dangers of AI interfering with elections and even a discussion about using AI to better understand the world’s oceans! The highlight was an incredible presentation by Will.I.Am from the Black Eyed Peas on AI and diversity. My children have played his songs over and over! I had no idea that such a talented musician can also speak about technology more eloquently than most data scientists that I know. He partnered with ElevenLabs to create an AI trained on data that represents communities like Compton and Watts. Key takeaways: ➡ AI algorithms are often controlled by a select few companies lacking minority representation. Will.I.Am is changing this narrative. ➡ He introduced Phylicia, an AI assistant designed to bridge gaps and make AI accessible to underrepresented communities. ➡ Phylicia explained complex concepts like quantum entanglement in relatable terms, highlighting its significance in AI's future. ➡ While billions are invested in making AI smarter than humans, Will.I.Am emphasized investing in the potential of diverse human minds. Check out this 7-minute sample to hear how Will.I.Am is creating a more representative AI that doesn't sound like an older British male. Will.I.Am's vision for inclusive AI is crucial as we shape our future. His presentation left me inspired by the potential of AI to create a more equitable world. #CogXFestival #WillIAm #AIRevolution #DiversityInAI #TechForGood
To view or add a comment, sign in
-
𝐂𝐨𝐮𝐥𝐝 𝐀𝐈 𝐛𝐞 𝐁𝐞𝐭𝐭𝐞𝐫 𝐚𝐭 𝐅𝐢𝐧𝐝𝐢𝐧𝐠 𝐃𝐚𝐭𝐚 𝐏𝐚𝐭𝐭𝐞𝐫𝐧𝐬 𝐭𝐡𝐚𝐧 𝐡𝐮𝐦𝐚𝐧𝐬? 🤔 AI can process information at an incredible rate. It can sift through millions of files and databases in seconds. Linköping University's research shows how AI, specifically their autoencoder models, can wade through complex biological data, shedding light on changes in our genetic code – an area where human analysis might take years. The LiU team's AI model managed to identify and comprehend patterns and detect not just the known genetic changes but also to unveil new ones, showcasing the sheer depth and efficiency of AI-driven data analysis. This is mind-blowing 🤯 . Keeping humans in the loop is vital to bringing invaluable intuition and context, ensuring that the data is used ethically, and protecting privacy. The blend of human expertise and AI's computational prowess will undoubtedly redefine our approach to data and help us make our world a better place. In 5 years, how do you see AI being used in data analysis? 👇 #AI #AIHealthcare #Data #Analysis #Future #Yobi
To view or add a comment, sign in
Ph.D. Student | Google CSRMP'23 | Fair & Responsible AI | AI Ethics | Data Scientist
1moGo Irene Solaiman!! 🤩