Collaborative Intelligence Week in Review - 02Aug2024
Best business article(s) I read this week…
Is Generative AI an Existential Threat to Human Creatives?
As the author writes, the answer is, “No.” He reasons using an analogy from financial economics developed by Nobel Prize-winning economist George Stiglitz. “If generative AI models can provide all the content humans need at low variable costs, then there is no incentive for humans to spend costly resources on content creation as they cannot profit from it. But if no human creates new content, then generative AI can only learn from stale information and be unable to generate up-to-date content that reflects new happenings in the physical world. This creates a paradox.”
Bridging the ‘Concept–Product’ gap in new product development
This is a long article. It focuses on FinTech subject matter experts and their efforts to incorporate AI into their products. The key takeaway is the authors’ development of a “contingency theory framework” and “process typology” that can help executives tailor AI innovation strategies to their organizations.
Best technical article(s) I read this week…
This is a short semi-technical paper describing the authors’ analysis of how well ChatGPT-4o and Gemini Flash perform when tasked with reading presentations with many charts and graphs. The authors found that “these models aren’t currently capable of reading a deck accurately end-to-end if it contains any complex or unlabeled charts. Even if a user created a deck of only labeled charts, the model would only be able to read 7-8 out of 15 labeled charts perfectly end-to-end.” Understanding the strengths and weaknesses of LLMs is helpful for managing expectations about them and for understanding where to incorporate them into existing workflows.
Other items I found valuable/interesting…
This AI-powered financial advisor has quickly gained $20 billion in assets
GlobalPrediction’s PortfolioPilot, an AI-powered investment advisor, has $20B in Assets Under Management and more than 22,000 clients. Is this an early warning for traditional asset managers?
This Scientific American article presents recent research showing that being polite to a Large Language Model produces higher quality output, but being excessively polite and overly flattering reduces output quality. An interesting look into the “minds” of chatbots.
Amazon develops a new AI model for its 'Just Walk Out' system
I’ve used Amazon’s “Just Walk Out” technology in airport stores. The company announced a new, multimodal AI foundation model that enables the system to “generate receipts faster, more accurately, and more efficiently.” Is this the future of retail?
Zoom CEO envisions AI deepfakes attending meetings in your place
This article is from June, but I just saw it this week. Zoom’s CEO announced that the company is pivoting to become an "AI-first company," with the goal of using AI to reduce human involvement in day-to-day work. One example cited is a digital twin “deepfake” that attends meetings for you and then summarizes the discussion and lists action items. While we’d all like to reduce the number of meetings we attend, do you think such an approach is viable? If so, would your digital twin be a valued helper or a threat to your job?
Coolest thing I saw…
Shape-Shifting ‘Transformer Bots’ Inspired by Origami
Researchers at North Carolina State University have developed a way to make a single plastic cubed structure transform into more than 1,000 configurations using only three active motors. The researchers will now focus on building more robust cubes that can handle higher loads.
A company that caught my eye…
Inceptive was founded in 2021 by Jakob Uszkoreit, one of the “Google Eight” authors of the research paper “Attention is All You Need,” which introduced the transformer architecture that powers today’s LLMs. Inceptive describes itself as a “biological software” company that is applying AI to drug development, with the goal of creating “medicine that is in much greater harmony with living systems than most existing medicines.”