The Ministry of Science and ICT has released the scores for the second stage evaluation of the 'Independent Artificial Intelligence (AI) Foundation Model' project. SK Telecom (SKT) achieved the highest score, while Motif Technology received the lowest. The score difference between SKT and Motif was 4.8 points.
On August 27, the ministry disclosed the detailed scores from the second stage evaluation of the project. SKT scored 70.6, followed by Upstage at 69.9, LG AI Research Institute at 69, and Motif at 65.8.
Earlier, Motif issued a statement requesting the release of detailed scores, stating, "It is necessary to explain why the benchmark conversion scores and expert evaluations yielded conflicting results so that the company and the industry can understand." The ministry responded that "other elite teams also agreed, and detailed scores will be made public."
The second stage evaluation consisted of three components: benchmark evaluation (AAII, NIA), expert evaluation, and user evaluation (AI experts and the general public).
In the AAII evaluation, Motif scored the highest with 11.9 points, followed by Upstage at 9.4, SKT at 8.8, and LG AI Research Institute at 7.8.
In the NIA evaluation, SKT's elite team ranked first with 13.4 points, followed closely by Upstage at 13.3, LG AI Research Institute at 12.8, and Motif at 12.7.
In the expert evaluation, LG AI Research Institute received the highest score of 29.5, with SKT at 29.3, Upstage at 29.1, and Motif at 27.1.
Among 49 representatives from AI startups participating in the user evaluation, SKT scored the highest at 11.6 points, followed by LG AI Research Institute at 11.3, Upstage at 10.8, and Motif at 8.4.
In the general public evaluation, which involved 185 participants, LG AI Research Institute ranked first, while SKT came in second with 7.5 points, Upstage third with 7.3 points, and Motif fourth with 5.7 points.
The Ministry of Science and ICT emphasized that the ranking results aim to drive qualitative growth and expansion of the domestic AI ecosystem, not merely to create a ranking competition or to eliminate teams. The ministry added, "We hope that following the release of these results, our companies will be evaluated not just by scores or rankings, but that our AI ecosystem will develop into one where new and competitive AI companies continuously challenge the status quo."
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.