Heading for the Holy Grail of AI

According to the MIT Tech Report, What’s next for AI in 2024, 2024 is seeing tech companies invest heavily in generative AI. Yet they will be under pressure to prove that they can monetize artificial intelligence. Google and OpenAI are going small. Both are developing user-friendly platforms that allow people to customize powerful language models and make their own mini chatbots. For the user, there are no coding skills required. Both companies have launched web-based tools to help users become generative-AI app developers.

This year generative AI might actually become useful for the regular, non-tech person. The MIT research indicates we’re going to see more people tinkering with a million little AI models. State-of-the-art AI models, such as GPT-4 and Gemini, are multimodal, meaning they can process not only text but images and even videos.

This new capability could unlock a world of new apps. For example, a real estate agent could upload text from previous listings, fine-tune a powerful model to generate similar text with just a click of a button, upload videos and photos of new listings, and simply ask the customized AI to generate a description of the property.

The AI Word at 3DExperience World

Here’s the day one recap of activities at 3DExperience World

At 3DExperience World, for the second year, AI was on everyone’s mind. We caught up with Suchit Jain, VP of strategy and business development at Dassault Systemes to see if his view was as positive this year as it was in 2023. Turns out he’s more bullish about the technology, but he is skeptical about how quickly we can put it to use.

He noted that it’s not easy to keep up with the pace of change in AI. “Technology moves faster than people can, so anything we can do to make technology easier matters,” Jain told Design News. “So much is happening around AI and generative AI.”

He explained that companies need to focus on giving AI a useful purpose. “You have to constrain the AI to make something useful,” said Jain. “The tools are getting better but they’re not quite to manufacturable. Right now, AI will be more assistive than generative.”

He also noted that the definition of AI was become broader in the past year. “Anything to do with data – on any computer – everybody calls it AI, said Jain. “For about 20 years now we have been working with data, data management, and big data. From that perspective we've been doing AI for a while. But if you talk about generative AI which creates something which didn't exist before, we’ve been doing AI for the last couple of years.”

Headed for the Holy Grail

For companies involved with PLM and design software, a major breakthrough arrived a few years ago with generative programs. These tools let the computer create an object from the purpose and constraints given by the user. These generative tools created many of the lattice shapes that deliver strength while reducing weight. This technology has found purchase in aerospace and automotive where reducing weight is imperative.

For many engineers generative design was the true beginning of artificial intelligence. “I first came in contact with generative design when it was creating complex shapes that often solved multiple problems,” said Jain.

Yet beyond generative design, beyond the chatbots, is the expectation that one day AI will create products from text. “The real Holy Grail in AI is going from text into a 3D product that is constrained and manufacturable.” We may be on our way, as a number of companies are already taking text into 2D drawings.

As for AI fear, Jain shrugged if off. “When the internet came out, everybody was worried about privacy,” said Jain. “Then the government and other people stepped in with laws and rules. That made the internet so much better. The same will happen with AI.”