AI Top-of-Mind for Nov 22
We’re taking a break Thu and Fri. Happy Thanksgiving!
‘Mike drop.’ Sam Altman back as OpenAI CEO. Additional coverage in the NY Times for subscribers.
Black Mirror or worse? AI drones and whether they need to be internationally regulated. From the DNI ‘Future of the Battlefield’ report:
More details on the ‘Artificial Intelligence Research, Innovation, and Accountability Act of 2023.’ The act addresses the following areas:
· Content Provenance and Emergence Detection Standards
· AI Definitions
· Generative AI Transparency
· NIST Recommendations to Agencies
· Risk Management Assessment and Reporting
· Critical-Impact AI Certification
· AI Consumer Education
A primer on the use of ReAct programming to enhance ChatGPT. From Andrea Valenzula in ‘ForCodeSake’:
In this article, I aim to provide a gentle introduction to ReAct prompting that comes integrated as part of the LangChain framework. This approach makes use of the chain-of-thought (CoT) prompting technique to enhance ChatGPT’s capabilities.
And you are interested in schooling up on Generative AI, Amazon is now offering a set of eight classes at no cost. Their goal is to ‘democratize’ AI and train up to two million people via their ‘AI Ready’ initiative.
Finally, more on AI hallucinations and how their parallel the human brain. From the ‘Medium’ article:
Our brains function on the basis of statistics, and from correlations of various kinds learn how to explain phenomena or make predictions. And as with algorithms, our brains sometimes hallucinate, not in the original meaning of the word, that of imagining things that do not exist or as a result of consuming psychotropic substances, but by attributing mistaken meanings to apparent correlations.
