"Concerns are growing that jobs will disappear as artificial intelligence (AI) spreads, but that is nothing more than needless worry. The tool called AI must be used in the right context to produce proper results, so humans ultimately have to be involved."
Suchit Jain, vice president of strategy and business development at Dassault Systèmes, met with reporters at the 3DExperience World 2026 event in Houston, Texas, on the 4th (local time), and said, "AI will handle repetitive tasks for us, but it cannot replace the fundamental engineering knowledge and know-how humans have built."
Dassault Systèmes, headquartered in France, is a software corporations leading 3D CAD (computer-aided design) and AI-based virtual twin technology. A virtual twin is a technology that replicates real factories and manufacturing processes in virtual space just as they are. It creates a "twin" of a real-world object in virtual space to run simulations, and uses that to predict results in advance or eliminate errors beforehand to optimize products and services. It counts global manufacturers such as Airbus, Boeing, Volkswagen and BMW among its major customers, and recently partnered with Nvidia to build "physical AI" for deployment on factory floors.
At this event, Dassault Systèmes unveiled three new AI agents embedded in its core design platform SolidWorks. The new AI agents, which Dassault Systèmes dubbed "virtual companions," support design work by taking requests typed into a chat window—much like conversing with ChatGPT—and drawing on various materials such as drawings and documents held by the operator.
Jain projected that as AI agents replace simple repetitive tasks, new types of jobs will emerge and the nature of work will change. Suchit said, "AI will not eliminate jobs; it will act as a copilot for engineers and creators, doubling productivity," adding, "Engineers and creators will not need to waste time on routine tasks and will focus on higher-level work."
He said the release of the new AI agents is part of Dassault Systèmes' effort to bring simulation—where it has a strength—into the mainstream of the design process, noting, "We are only now seeing AI truly take shape across design and manufacturing," expressing optimism. However, he added that it is still not at a stage where one-line prompts can yield detailed, immediately manufacturable blueprints.
He said, "You cannot get a perfect design without engineering-level precise detail," and added, "If you simply tell AI, 'Design a smartphone,' it will produce an output, but to get the design we want, you need an engineer's specific and precise guidance." The point was that one must know how to use the tool called AI in the right context based on expert insight and acumen.
In a similar vein, he also dismissed concerns that AI will replace software. At this event, Dassault Systèmes executives predicted that as the number of AI agents assisting engineers grows, the software and tools they use will increase exponentially. Jain said, "AI will only make the engineering process more productive; it will not replace jobs or software."
As for why the AI agents were released in three—"AURA," "LEO" and "MARIE"—he said, "Because the required compute varies by the agent's functions, we introduced three to optimize compute expense and efficiency," adding, "If you insist on a single system, you may end up running an unnecessarily complex and expensive system even for a simple question."
Asked about Dassault Systèmes' role in the robotics market, considered a core industry of the future, he said, "We do not directly develop vision detection software for Humanoid Robot, but we can receive signals generated by that software and influence the robot's mechanical design." He added, "We will provide the key connection points in the process of developing the robot's hardware, mechanical systems and electrical systems."
Jain said Korean manufacturers and startups have a relatively high level of AI adoption among major countries. He said, "If we divide the Korean market into three stages—'indifferent,' 'aware but not using,' and 'aware and actively using'—Korea is clearly in the 'actively using' stage," adding, "I cannot be sure whether Korea itself is leading the development of next-generation large language models (LLM), but what matters is that it chooses the right tools and has a firm grasp of every technology that can be used on the manufacturing floor."