AI Weekly: Digging into DALL-E 2 for the enterprise

Welcome to the latest edition of the AI Weekly newsletter 

As I finish my first week at VentureBeat, it’s a perfect opportunity to introduce myself: I’m Sharon Goldman, a senior editor and writer covering AI for technology decision-makers. 

Based in central New Jersey (exit 10), I’ve reported on business-to-business (B2B) technology for over a decade, writing for publications including CIO, Forbes.com, Insider, Shopper Marketing, Adweek and CMSWire.

I chose a busy AI news cycle to get on board at VentureBeat. Certainly, the debut of DALL-E 2, OpenAI’s new AI model, which uses advanced deep-learning techniques to generate and edit photorealistic images simply by comprehending text instructions, has been the subject of chatter for two weeks now. That includes both rhapsodic responses around DALL-E 2’s capability to create amazing photos of avocado-shaped teapots and chairs, as well as loud concerns about possible digital fakes image generation and the spread of misinformation. 

As Ben Dickson explained here, “DALL-E 2 is a ‘generative model,’ a special branch of machine learning that creates complex output instead of performing prediction or classification tasks on input data. You provide DALL-E 2 with a text description, and it generates an image that fits the description.”

What sets DALL-E 2 apart from other generative models, he continued, is “its capability to maintain semantic consistency in the images it creates.” I wanted to know what this all means for enterprise business, so I reached out for comments from a couple of experts: 

Finally, in a VentureBeat column this week, Sahor Mor, a product manager at Stripe, explored how DALL-E 2’s powerful text-to-image model might be used to generate datasets to solve computer vision’s biggest challenges. 

“Computer vision AI applications can vary from detecting benign tumors in CT scans to enabling self-driving cars, yet what is common to all is the need for abundant data,” Mor wrote. “DALL-E 2 is yet another exciting research result from OpenAI that opens the door to new kinds of applications. Generating huge datasets to address one of computer vision’s biggest bottlenecks – data – is just one example.” 

Some experts, however, maintain there is the danger of over-hyping DALL-E 2. “It’s important not to conflate the ability to generate realistic images from text with “understanding,” Peter Stone, president, founder and director of the Learning Agents Research Group (LARG) within the AI Laboratory in the department of computer science at the University of Texas at Austin, told VentureBeat. “I do not think of DALLE-2 as making significant advances (beyond existing models) towards the long-term goals of many people in the field of AI – it does not give me any more confidence than I had before that all of AI can be solved with neural networks alone.” 

In Case You Missed It

AIs future is packed with promise and potential pitfalls
Why it’s a must-read: Solving the inherent problems of foundation models requires real-world use.

7 ways to improve data for supply chain digital twins
Why it’s a must-read: Various approaches to supply chain twins show tremendous value in sorting out supply chain bottlenecks, improving efficiency and meeting sustainability goals.

The success of AI lies in the infrastructure
Why it’s a must-read: AI is all about data, and data lives in infrastructure. The only way to ensure that AI’s promises can be turned into reality is to create the right physical underpinnings to allow the intelligence to work its magic.

Can human-centered MLops help AI live up to hype?
Why it’s a must-read: Human-centered AI is more than a hyped buzzword or philosophical framework. While it focuses on how AI can amplify and enhance human performance, it is really about helping enterprises build and manage better AI.

Thanks for reading,

Sharon

Twitter: @sharongoldman