Chief Scientist & SVP of AI @ Dataminr. Scientist, engineer, innovator, artist, machine learning, AI. Ex Yahoo, IBM, FujiXerox, KAIST, Columbia U. Ph.D.
We're sponsoring CVPR this year. Our platform ingests data from hundreds of thousands of public data sources from all over the world, applies multi-modal techniques to detect events, and uses generative AI to summarize those events for emergency response,…https://t.co/WHYwwUtJiG
We have very few spots left at the hackathon on LLMs for Good at Cornell Tech next week, June 8th (Cornell Tech Fest). In addition to the hackathon, a great lineup of speakers! #llms#aiforgood#ai https://t.co/G3nt6CfGPp
I'm organizing a hackathon on LLMs for Good at Cornell Tech next week, June 8th as part of Cornell Tech Fest. Limited spots- in addition to the hackathon, a great lineup of speakers! #llms#aiforgood#ai https://t.co/G3nt6CfGPp
Interesting paper by Eleftheria Briakou (one of our previous interns) on "incidental bilingualism—the unintentional consumption of bilingual signals, including translation examples" in LLMs. https://t.co/e5qMuEexpP
One of my Deep Learning students at Cornell Tech, @kanlanc is co-organizing a GPT Hackathon this weekend in NYC. IF you're a student in the NY area, you're highly encouraged to apply! $50K in cloud credits, Claude beta access & var…https://t.co/5V6zhhHZfg https://t.co/slG2891sRV
Interesting new model from Meta (open source). As I've said before, LLMs are really LXMs, where X is any kind of data (here, images, audio, text)- gives a glimpse of what's coming.. https://t.co/agizzGi7Db
Nice to have our AI Org's work featured in this Amazon Web Services (AWS) customer story- one of the keys to our success as a company is having an amazing team of highly skilled AI Engineers working with our scientists to scale pro…https://t.co/FFaPKuwloj https://t.co/Qnp2ujqZBM
Potentially impactful paper using chain-of-thought reasoning to train smaller models (outperforming LLMs with models that are 2000x smaller). https://t.co/m9lofwPPil
Very much looking forward to speaking tomorrow at PeacCon 2023, alongside @SamPowers1, on The Role of Emerging Technologies in Local and Global Peacebuilding !
#aiforgood#llms https://t.co/UyMxDwBsBQ
We just launched a new program. If you are an NGO or multilateral organization, the Social Innovation Lab at Dataminr and my org (AI) would love to partner with you on building AI solutions to drive transformational impact for people and the planet. https://t.co/7D8yYRgmXa
Text generation in images has been a challenge for AI models. DeepFloyd IF, a new model by Stability AI l, released today, seems to significantly improve the state of the art. And it will be released as open source. https://t.co/lGE0vyqw9B
As GenAI adoption grows, so will cyber risks. Proud of my org (and the Dataminr team) for our recent Cyber product launch. Building this at scale effectively requires lots of Deep Learning (LLMs++), domain expertise, lots of data,…https://t.co/1PjUF6Hrfe https://t.co/dh7fki1mNm
We’re hiring Ph.D. Students for internships this summer. Large Language Models (LLMs) ++ :)
Apply below if you want to do challenging AI work that has a positive impact on the world.
https://t.co/WIou6e0G6J
My friend Foster Provost is performing next week at The Bitter End. I created the video for the single "Remember Me" (video release next week, sneak preview on Spotify). And guess what? I didn't use AI (unless consider DaVinci Resolve an AI tool) :) I'll…https://t.co/YRakaxY5CW
@JournalistJill Had a similar experience w my son when he was younger. One day he actually lost his stuffed animal. We told him it had gone on vacation. Years later he saw a picture of himself with it and said “oh that’s the doggy I lost and you told me it had gone on vacation”.
Drago was a friend, and a great colleague- we worked together at Columbia and many years later at Yahoo (where we published our last paper together). We stayed in touch over the years. I saw Drago recently- as always, he was full of life & ideas; optimist…https://t.co/OewmJ2Pk96
A couple of weeks ago I participated in a panel on GenAI (along with Vijay, @gkm1, @xdh, and @joshim5. Vijay K Narayanan did a great job summarizing our discussion.
Opportunities (and challenges) abound and two factors will be particularly critical: spe…https://t.co/dSNHsPMhC2