Suggestions
Paul Greenberg
Executive with major impacts in digital media and AI applications
Paul Greenberg is the CEO of Butter Works, a digital consulting firm specializing in using predictive AI for video content creation and distribution. He has a notable background in digital media and has held several executive roles throughout his career.
Education
Greenberg graduated from St. Albans School in Washington, D.C., in 1986 and earned a bachelor's degree from Columbia College in 1990. He later received an MBA from Columbia Business School in 1997.
Career Highlights
-
Butter Works (2018 - Present): As CEO, he implemented advanced AI techniques to optimize video production and distribution strategies, serving clients including Fortune 50 companies.
-
Meet Cute (2022 - 2023): He served as President and COO, utilizing generative AI to enhance video production efficiency significantly.
-
A+E Networks (2016 - 2018): He was the Executive Vice President and General Manager, where he led the digital video studio and achieved substantial audience growth.
-
Previous Roles: Greenberg has also been CEO of CollegeHumor and Nylon, and held senior positions at Time Inc. and MTV Networks, where he was recognized for his contributions to digital media and audience engagement.
His extensive experience in digital strategy and media management has established him as a prominent figure in the industry.123
Highlights
We’re one step closer to cracking the code of how AI thinks.
As impressive as generative AI chatbots are, even their creators don’t know exactly how they come up with the responses they do.
Researchers at Anthropic (parent company of the AI model Claude) released a paper today saying they’ve made significant progress toward reverse engineering their large language model to understand why it comes up with specific outputs.
Using a technique called “dictionary learning,” the team was able to associate a combination of neurons that, when fired in unison, evoked a specific concept.
They can now tune this feature to some extent to dial up or down certain characteristics of the model with the ultimate goal of making it safer to use and demonstrating a lot less bias.
Dictionary learning still can’t break open the AI black box, but it gives us an important tool to use to start to peer inside.
Do you think we'll be able to figure all of this out soon, or will gen AI remain a mystery?
#ai #llm #largelanguagemodel #anthropic
I'm speaking on a panel about AI this coming Wednesday at 11 AM in New York City. Should be a great discussion (and it's free). Please join Paul Melcher, Francis (Frank) Evers, Axel Ebermann, and me. Here's the info: https://lnkd.in/e4mbVwHt
