TEDxNaperville的动态

Here, by popular request, is the entire performance. What a superb audience! https://lnkd.in/e2gvFuv5 Nota Bene: No, I won't rewrite my message with AI! How cheeky to asked an English teacher...Obviously, linkedin isn't using AI to determine whom to ask for AI to rewrite a post. Then again, perhaps I'll do so next time and see how many more people react to AI's personality.

I spoke about the new pyramid builders: AI, wealth, and the future of work at a TEDx by Aalto University today. It's about the rise and fall of civilizations, and what today's techies can learn from the ancients. Its three sections, in short: 1?? On the spring equinox, the Pyramid of Kukulkhan in Chichen Itza seems to reveal a "descending serpent"—which could symbolize a real cosmic phenomenon. The Maya devoted vast resources to meticulously study the skies and collect data over centuries, and build accurate calendars, symbolizing humanity’s quest to accurately predict important events. 2?? Now, Google’s Waymo and AI models like OpenAI's GPT and Anthropic’s Claude are modern "prediction machines." Like the Maya in their day, companies are investing billions into systems that precisely predict the future based on immense amounts of data—whether self-driving cars or even AI-run laboratories. 3?? AI is transforming jobs once handled by humans, leading to massive economic shifts. Will everyone benefit, or could we see wealth concentrated among a few tech giants, creating inequality reminiscent of the Maya’s downfall? History hints that unchecked imbalance can lead to societal collapse. Will this time be different?

?? Another big win for open source AI from France! ???? Following the success of Mistral AI, Kyutai has unveiled a real-time native multimodal foundation model trained from scratch. Kyutai showcased its real-time native multimodal foundation model trained from scratch. Its a privately funded non-profit French AI research lab that attracts PhD students, postdocs and researchers and promises to publish its research and open source its foundational models. As open-sourcing catches up with closed models at an unprecedented rate (access to NVIDIA chips is no longer a bottleneck), closed models make less sense given the trade-off of not advancing research and lack of transparency. My thesis, supported by developments such as more user features in ChatGPT and artefacts in Anthropic, is that as LLMs become commoditised, providers of closed source models will focus on product, while frontier research will increasingly be driven by open source. ???? This shift will accelerate the realization of AI’s economic potential and hasten the path to achieving AGI. ???? #AI #genai #opensourse #AGI

https://lnkd.in/eVjtUM-3 https://lnkd.in/eNQiEGRy Join MIT associate professor Justin Reich in this entertaining podcast on teaching in an era of generative AI.?He?walks?us through the inner workings of systems like ChatGPT and?addresses?the challenges educators face when students use generative AI to take the easy route with homework and bypass cognition. If you don't have an hour to commit, scroll to 27 minutes and listen to his account of MIT's "Beaver Rush." It's a generative AI hallucination, and it's hilarious! #educators #education #generativeai #edtech #k12education

?? New white paper: Validating predictive AI in systematic review screening Our validation study showcases how machine learning with predictive scoring enhances systematic review quality and efficiency: ?? AI accuracy: 87.4% (high confidence scores >0.95 predicted inclusion) ??Traditional screening accuracy: 86.7% ??94.8% inter-rater reliability ??Complete audit trail & human oversight ??Built on machine learning, not generative AI ??Screening time: 2-3 weeks → 2-3 days Our predictive AI identifies high-probability includes with greater accuracy than traditional methods while maintaining full EMA and NICE compliance. Read the white paper to see how our validated AI process transforms evidence generation: https://lnkd.in/eh74A3b9 #RegulatoryCompliance #SystematicReviews #EvidenceGeneration #marketaccess

The Artificial Intelligence Podcast featured Michael Huth, Chief Research Officer and Co-Founder of Xayn, to share his insight on AI legislation, the integration of specialized AI, and his journey in the industry. “At the end of the day, AI products want to be sold around the world, so we have to make sure that there is enough alignment to actually have the economic growth that AI might enable,” shares Huth. What challenges does AI face in highly regulated areas? Listen here: https://lnkd.in/ex-Jst_K #SlicedBrand #TechPRAgency #Techpr #publicrelations #PRAgency #AI #AIregulation #Xayn #artificialintelligence

How can we use AI to construct a virtual laboratory that simulates policy outcomes for S&T innovation and competitiveness? https://lnkd.in/djkm_d2E

We've run a validation study on our AI-powered approach vs traditional methods and how this ties into guidance from the EMA and NICE. Reflecting on the feedback on the use of AI at #ISPOR , we're confident that we can still optimise and streamline processes while keeping on the right side of HTA bodies #RegulatoryCompliance #SystematicReviews #EvidenceGeneration #marketaccess

Interested in how #smart environments can support #aging-in-place, you are wondering how to guarantee #humanistic values with #AI integration? This article is for you! Thank you Amy Hwang, PhD and Rosalie H. Wang for the leadership, and the whole team for the exciting exchanges that led to this publication.