Plainsight Blog

AI Week in Review #29 | 2022

July 22, 2022
Bennett Glace

Bennett Glace

This week in AI & Machine Learning: The next stage of development for DALL-E 2, ML for particle accelerators, and more. 

A Note from the Author 

Industries like retail and dining depend on service excellence. The customer is always right and they’ll often take their business elsewhere when the shopping or dining experience is lacking. In honor of Get to Know Your Customers Day (a quarterly celebration), we looked at some of the ways computer vision can help enterprises generate customer insights to set themselves apart and deliver a higher caliber of customer service. Check out our latest blog to learn more.  

Artificial Intelligence News

DALL-E 2 to Enter Beta Stage

OpenAI’s DALL-E 2 has taken the internet by storm since debuting earlier this year. Prompted with short descriptions from users, the AI model is capable of creating high-quality “photographs” and even mimicking the style of well-known artists. Experts believe it is one of the most advanced tools of its kind, but only a select group of users have had access thus far. 

DALL-E 2 is not to be confused with DALL-E mini, a free imitator that recently changed its name to Craiyon. Though it’s unclear whether or not OpenAI’s more powerful tool will ever be available to the general public, anyone can try out Craiyon with their own prompts today. 

Buzz over DALL-E 2 quickly inspired people from all over to join a waitlist.  On Wednesday, OpenAI announced that they’ll offer Dall-E 2 access to one million individuals selected from this list over the next several weeks. Users selected to participate in the beta will have the option to create 50 free images during their first month and an additional 15 free images throughout subsequent months. Learn more about what broader access to DALL-E 2 means, how OpenAI is working to define content guidelines, and what the future could have in store.  

AI-Generated Images from Meta

DALL-E 2 may be impressive, but the team at Meta would argue there’s no guarantee it will accurately reflect a user’s instructions. Imagine asking it to create an image of a person riding a bicycle. While this may sound like a simple, predictable composition, there are in fact a number of variables: Which direction is the rider facing? How large is the bicycle? Is the rider wearing a helmet? 

Meta believes its new solution, Make-A-Scene, allows for greater specificity and creativity in AI-generated images. Making use of both sketches and text prompts, Make-a-Scene has been shown to accurately reflect the user’s vision and Meta hopes it will soon empower artists to express themselves effectively across a wide range of media. View a gallery of images created using Make-a-Scene and read more about Meta’s ongoing efforts to improve AI-generated art.   

Improving Particle Acceleration with Machine Learning

New insights into the subatomic world could soon become available thanks to a machine learning platform capable of controlling particle accelerators and lasers more effectively than traditional approaches. Research led by members of the Berkeley Accelerator Controls and Instrumentation (BACI) program has helped create tools capable of responding to trouble, learning from mistakes, and adapting their strategies to better stabilize beams and drive subatomic studies forward. 

Next steps for the researchers include trying out new machine learning on edge computing devices. Read a summary of ongoing initiatives that could provide for the next generation of subatomic research and serve a number of medical, industrial, and scientific applications. 

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