Plainsight Blog

AI Week in Review #30 | 2022

July 29, 2022
Bennett Glace

Bennett Glace

This week in AI & Machine Learning: An AI to predict the shape of every imaginable protein, Google’s new & improved digital assistant, and more. 

A Note from the Author  

The dog days of summer are often considered the doldrums for sports fans with just late-season MLB action available on TV. Fortunately, when computer vision and sports come together, there’s always exciting research and initiatives promising to change the on-field and in-stadium experience. Tide yourself over between now and the return of your favorite team by checking out our blogs on the ways vision AI is transforming baseball analytics and providing for the next generation of stadiums

Artificial Intelligence News

An AI-Powered Database of Every Single Protein

Thanks to the revolutionary AI network known as AlphaFold, scientists can now assess the 3D shape of every known protein in the universe as easily as they might conduct a Google search. The lab behind AlphaFold, DeepMind, began developing the technology back in 2020 and released predictions for more than 300,000 proteins last year. Thursday, July 28th, they added hundreds of millions of additional predictions to the database. 

Prior to DeepMind’s innovations, determining a protein’s shape required intensive testing. The DeepMind team hopes their technology will not only contribute to speed up existing research, but potentially give birth to an entirely new field of study: metaproteomics. 

Speaking to The New York Times, DeepMind’s CEO, Demis Hassabis, described just some of what AlphaFold could enable. “Scientists can now explore this entire database and look for patterns,” he remarked, “correlations between species and evolutionary patterns that might not have been evident until now.” Experts are also hopeful that the technology will help a new generation of scientists deeply immerse themselves in the world of structural biology more quickly than their predecessors could’ve imagined. 

Google’s ‘Look and Talk’ Paves and the Future of Digital Assistants

Digital assistants like Amazon Alexa, Siri from Apple, or the Google Assistant generally rely on prompts known as “hotwords” to begin conversations. Users issue commands and ask questions by first saying, “Hey Google” or a similar phrase. This is convenient enough for users, but it’s hardly a close analogue of real conversations between people. Gestures and glances are just as likely to kick-start conversations as the verbal cues digital assistants have traditionally relied on. 

Leveraging a combination of 8 different machine learning models, Look and Talk’s advanced algorithm can differentiate between passing glances and meaningful ones and engage users in conversation from five feet away. Learn more about how Google developed Look and Talk, some of the challenges engineers overcame along the way, lingering obstacles, and the precautions Google is taking to protect user privacy. 

Can AI Help Congress Grow More Effective? 

On Thursday, July 28th, the Select Committee on the Modernization of Congress (established in 2019) held a hearing in the House of Representatives. Topics of discussion included the ways AI and machine learning may be able to transform the country’s legislative body. The conversation focused partially on how AI could help representatives better understand how existing laws are perceived as well as the potential impacts of new proposals. 

AI could also serve a simpler, more logistical purpose for members of Congress. An average congressional representative serves on a number of committees and subcommittees. This can make scheduling meetings and hearings an ordeal, but AI has the potential to help sort things out. Read more on what was discussed during this week’s hearings and the ongoing efforts of the Select Committee. 

Join our Community

See you next week! Until then, keep the conversation going on Plainsight’s AI 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 project strategy, through model deployment, and ongoing monitoring, Plainsight helps customers successfully operationalize vision AI applications to solve highly diverse business challenges.

Share