AI NewsWaveOct. 31 AI Briefing: From Samsung Executives' Meeting With Sam Altman to AI Investment Jitters
Key points
- Samsung executives and SK chairman each meet Altman
- Worries over AI investment returns; OpenAI also trims its fundraising target
- Google runs AI inference on a satellite in space; OpenAI restricts teen use
About 50 Samsung executives held a talk on the 31st at the Samsung Human Resources Development Center in Yongin, Gyeonggi Province, to which they invited OpenAI CEO Sam Altman, and discussed management strategy for the AI era. AI Newspaper (인공지능신문) reported that they discussed how AI will change the industrial landscape and management strategy for the next 10 years. The same day, according to business circles, SK Group Chairman Chey Tae-won also had lunch with Altman in Seoul, and key executives of SK hynix and SK Telecom were reported to have attended.
Chosun Biz, citing the Financial Times, reported that OpenAI cut its external fundraising target by $10 billion. EBN reported that the target of $30 billion last month became $20 billion. Concern over AI investment continued in the market. According to IT Chosun, Intel shares plunged 28% following Alphabet, and AI Times said in its headline that $2 trillion in U.S. tech market capitalization evaporated over the month of October.
Amid this, Apple regained the No. 1 spot in global market capitalization from Nvidia for the first time in three months at the close of New York trading on the 30th (local time) (IT Chosun). Kukmin Ilbo reported that Apple, which had stepped back from the AI investment race, emerged as a relatively safe investment.
AI Times reported that Google ran TPUs, data-center-class AI chips, on a satellite in Earth orbit and succeeded in sending AI inference results to the ground. iNews24, citing Yonhap News, reported that Google said it received the first AI response on the 30th (local time) from a prototype satellite for Project Suncatcher, its space data center plan.
According to AI Times, OpenAI will restrict weapons-related information in ChatGPT to prevent violence by teenagers. Even for questions that show no violent intent, it will not give teenagers specific information on weapons and security systems, and it will automatically end the conversation if an obsession with violence is detected.
According to AI Times, Anthropic will rename its highest-priced Claude plan and remove the five-hour usage limit for verified subscribers. The explanation is that some users took the 20x shown on the plan as a multiplier that also applies directly to weekly usage, causing confusion, so the company renamed the plan and explained its usage terms.