"AI handles the coding, developers focus on service perfection"... 28 hours for 100 LG developers

In an era where artificial intelligence (AI) performs coding, how is the role of the developer changing? The '2026 LG Promptathon Challenge' was held, bringing together approximately 100 developers from LG Group to create actual service-level results using AI models. Unlike traditional hackathons, this competition focused on increasing the perfection of services within a given time by utilizing LG's AI model, 'KX1', instead of developers coding directly.

"Service planning and inspection instead of coding"... A changing development environment

During this challenge, which lasted 28 hours, the working methods of developers showed a starkly different appearance from the past. According to an LG Group video, while past hackathons were dominated by developers spending time typing and coding directly, this competition saw a higher proportion of issuing commands to AI and inspecting the resulting outputs. A developer participating on-site shared the change in the field, saying, "Last year, because the performance of AI tools was low, there was a lot of sound from direct coding, but now, because AI performs the development if you just give a command, the sound of keyboards has decreased and the sound of conversation is heard more often."

Development using AI showed a huge difference in terms of productivity. Participants in the video showed positive reactions, saying, "It would have been impossible to do this much work in a single day, but by working with AI, we can unfold more of what we think about." However, as the importance of 'inspection'—where humans must directly check and verify the results generated by AI—grows, observations have emerged that the developer's work is shifting from 'control' to 'following the AI.'

Replacing tens of millions of won in equipment with 500,000 won range... Practical ideas for the field pour in

Developers from various LG Group affiliates participating in this competition presented practical ideas that could solve problems in the field. In particular, projects aimed at cost reduction and work efficiency drew attention. One team presented specific economic effects by planning a support tool that could replace expensive vehicle system development equipment worth tens of millions of won with equipment in the 500,000 won range.

Ideas specialized for the home appliance field also emerged. An agent service where users can arrange home appliances or receive product recommendations based on house floor plans was developed, along with services that use AI to drastically reduce the amount of data required for robot learning. In the case of robot learning, a technical approach was attempted to shorten a process that previously had to be performed at least 200 times down to about 10 times through AI. Additionally, attempts to solve the 'Pain Points' of the field continued, such as platforms that help sales representatives easily grasp hardware and software specifications to recommend them to customers.

Concerns in the AI era, "Between productivity improvement and role change"

The productivity revolution brought by AI brought both expectations and fundamental concerns to developers. A junior developer appearing in the video mentioned the situation where AI allows one person to do the work of ten, saying, "Some people think that since one person does the work of ten, the remaining nine are not needed. I am also scared because I don't know which direction my work will flow, but everyone agrees that AI increases productivity."

The competition proceeded in a manner where teams that passed the preliminary rounds through a fierce 28-hour process underwent a 4-week advancement process before taking the final stage. Officials from LG AI Research focused on evaluating whether the services implemented by the developers at the site had enough perfection to be introduced into actual field operations, and how effectively they utilized the KX1 model.