This week in AI & Machine Learning: Precision livestock counting, a debate about sentient AI, and recommendations from the FTC.
A Note from the Author
Everybody’s talking about Plainsight! This demonstration of our Precision Livestock Counting model is having a moment on LinkedIn, earning thousands of shares over the last week. Check it out for yourself and visit our blog to learn more about how we made it work. While you’re there, read up on additional vision AI use cases for agribusinesses and Plainsight’s exciting collaboration with one of the world’s largest food processors. As our partner NVIDIA highlights in their post, Plainsight deployed models for not only counting animals, but monitoring their health and well-being with real-time insights too.
Artificial Intelligence News
Does Google’s LaMDA AI Have a Soul?
On June 11th, Google engineer Blake Lemoine published a transcript of an ‘interview’ he and a colleague conducted with the company’s AI chatbot. Lemoine had been tasked with determining whether or not the Language Model for Dialogue Applications showed biases in its interactions with users. Published to Medium, the conversation became an instant viral hit and Lemoine soon found himself profiled in The Washington Post and placed on administrative leave by Google.
His interview subject has become the source of much speculation throughout the last week, with experts across the AI sector weighing in. Lemoine believes LaMDA’s discussions of loneliness, a desire for knowledge, and the solace found in meditation suggest not merely an advanced technological brain, but a soul. Most experts have rejected Lemoine’s conclusions, including his employer. Speaking to NPR, Lemoine describes being “laughed out of the office” and a spokesperson from Google deflated speculation about sentient AI in an official statement. “Of course,” the statement reads, “some in the broader AI community are considering the long-term possibility of sentient or general AI, but it doesn’t make sense to do so by anthropomorphizing today’s conversational models, which are not sentient.” Read NPR’s summary of the ongoing conversation and several dissenting perspectives.
The FTC Encourages Congress to Approach AI with Caution
In 2021, Congress commissioned a Federal Trade Commission (FTC) investigation into the ways AI can help address a number of persistent “online harms,” including fraud, impersonation, illegal drug sales, and media manipulation. On Thursday, June 16th, the FTC delivered a report encouraging members of Congress not to approach AI as a technological cure-all. They should instead, the report advises, use “great caution” in attempting to leverage AI for reaching policy decisions.
In particular, the FTC warns about the ways built-in biases and other design flaws can diminish AI’s efficacy, as well as the ways AI often encourages reckless data extraction. Rather than finding strategies for AI to combat online harms, the FTC suggests introducing legislation to safeguard against additional dangers. Ultimately, the FTC decided to send its report to Congress by a 4-1 vote.
AI for Recognizing Cross-Linguistic Signs of Autism
Among the key indicators of Autism Spectrum Disorder (ASD) are differences in speech prosody, which refers to the modulation of the rhythm and tone of one’s voice to communicate emotion and intent. Because even subtle deviations from typical speech prosody can significantly impact socialization, understanding them is considered an important part of addressing ASD.
Numerous studies have helped recognize the prosody differences associated with ASD, but the underlying causes are not well understood. What’s more, these studies have tended to focus on English-language speaking groups alone. New research conducted by teams in Hong Kong and at Northwestern University and published in PLOS One explores differences in speech prosody across English and Cantonese speakers with and without autism. Though the languages are about as distinct from one another as possible, the researchers observed certain commonalities. They believe their research could pave the way for establishing prosodic profiles on additional languages and better understanding autism in general. Learn more in this write-up from Northwestern.
Join our Community
See you next week! Until then, keep the conversation going on Plainsight’s AI for All Slack Channel.
About the Author & Plainsight
Bennett Glace is a B2B technology content writer and cinephile from Philadelphia. He helps Plainsight in its mission to make vision AI accessible to entire enterprise teams.
Plainsight’s vision AI platform streamlines and optimizes the full computer vision lifecycle. From data annotation through deployment, customers can quickly create and successfully operationalize their own vision AI applications to solve highly diverse business challenges.