I am pretty surprised by the lack of urgency among many (but not all) academics in addressing what seems to be one of the biggest developments in modern times – we accidentally built a machine that produces something that looks like language & thought Why? What does it teach us?
@emollick
-
Beyond Cost Cutting: Strategic AI Implementation for Organizations
By
–
One concern I have about AI implementation is that business leaders will see it purely as a way of cutting costs, because of a failure of imagination. This will hurt companies As I discuss in this Forbes interview, organizations need to think more deeply. https://
forbes.com/sites/jenamcgr
egor/2024/03/25/your-employees-are-your-rd-lab-how-ai-changes-what-matters-at-work/?sh=2b96fd6326cb
… -
New AI Book Co-Intelligence Released with GPT Tools
By
–
Very honored that my new book on AI, Co-Intelligence, out April 2, is recommended by @AdamMGrant in his newsletter. Plus, if you pre-order it, you can get access to a bunch of GPT tools to help you explore the the book (see my pinned tweet for details). https://
adamgrant.substack.com/p/the-8-new-id
ea-books-to-start-spring?r=i5f7&utm_campaign=post&utm_medium=web&triedRedirect=true
… -
IT and Worker Views Diverge on AI Implementation Strategy
By
–
Different "thought worlds" are a problem in AI. The IT teams in charge of AI think of "implementation" as scale: worrying tokens cost a lot, inference is slow, customization needed. The workers using AI often view "implementation" as "let me use ChatGPT how I want 20 times/day."
-
Benchmarking AI Systems Against Human Performance Standards
By
–
To me, the missing need is good comparisons to humans. But that is a complex and hard standard as well.
-
GPT-4 Performance Analysis: Consultants’ Diagram Summarization Accuracy
By
–
Claude 3 on a diagram summarizing our paper on the performance of consultants using GPT-4 (if you don't know it: https://
oneusefulthing.org/p/centaurs-and
-cyborgs-on-the-jagged
…) On one hand, it nails most of it, very impressive. On the other, it makes one error that took me a bit to spot So is this a good or bad job? -
LLMs as Laboratory for Story Transmission Across Generations
By
–
A cool use of LLMs as a lab for social science: what can we learn about the transmission of stories across time from AI? This research effort got LLMs to pass stories through 50 “generations” altering the personalities of the storytellers. The results echoed human story patterns
-
GPT-3.5 Excels at Summarizing Medical Papers
By
–
A useful finding: many people use AI to summarize technical articles but we have little evidence of how good it is at that. This paper shows that GPT-3.5 (not even 4) does a good job on medical papers Doctors rate the median quality as 90/100, accuracy as 93/100 & bias as 0/100.
-
AI Reliability Varies: Prompting Experience Creates Perception Divides
By
–
A big problem with AI is if you don't want it to work for something it won't work. If you want it to work for something, you often (but not always) can make it work with some prompting and experience. It leads to stark divisions in how people see AI.
-
ChatGPT-4 Standalone: No Azure, No Custom Setup Required
By
–
Just ChatGPT-4, no Azure, no special tuning, no custom apps.