[주간 클라우드 동향/11월④] The era of ‘AI agents’ heralded by MS

e-biz*solution

[주간 클라우드 동향/11월④] The era of ‘AI agents’ heralded by MS

Digital Daily
Publication date 2024-11-25 10:17:08

[디지털데일리 권하영기자] Industry interest increased as Microsoft (MS) released a number of new services last week. At ‘MS Ignite 2024’, attended by over 200,000 people from around the world in Chicago, Illinois, USA from the 19th to the 22nd (local time), MS showed off its competitiveness in cloud-based generative artificial intelligence (AI) technology.

According to the results of a study conducted by Microsoft on about 4,000 business leaders around the world through IDC, the adoption rate of generative AI jumped from 55% to 75% in just one year. Companies that introduced AI recorded an average return on investment of 3.7 times, and in particular, leading AI companies achieved a return on investment (ROI) of 10.3 times.

In his keynote speech, CEO Satya Nadella emphasized, “AI is doubling its performance every six months,” adding, “This is a rate that exceeds Moore’s Law, which doubled performance every 18 months.” . However, CEO Nadella mentioned the debate that the ‘AI scaling law’ (proportionality of AI model size and performance) has reached its limit and analyzed, “This debate will lead to innovation in model architecture, data system, and system architecture.”

In fact, MS is leading changes in daily work by innovating ‘Co-Pilot’ at every moment. On this day, MS also unveiled a large number of ‘AI agents’ that increase the business value of Co-Pilot. ‘SharePoint Agent’, which provides document-based insights by linking an AI agent to a specific site or file, ‘Interpretation Agent’, which provides real-time interpretation in 9 languages ​​in Teams, and ‘Project’, which automates the entire project from planning to progress management. ‘Manager Agent’, etc.

In this way, MS appears to be opening the era of ‘AI agents’ and incorporating generative AI services more deeply into our daily lives. In particular, there appears to be a move to further accelerate the pace by actively promoting the fact that the introduction of AI is ‘profitable’. This is evident just by looking at the introduction of the ‘Copilot Analytics’ function, which allows companies to directly measure the ROI of AI introduction for each department such as sales, marketing, and finance, and visually see the impact on business performance.

At this event, global IT companies strengthening their linkage with Microsoft also stood out. Open source company Red Hat decided to strengthen integration with MS Azure to simplify generative AI deployment, and remote work solution company TeamViewer unveiled ‘AI Session Insight’, which integrates MS Teams, Co-Pilot and Azure Open AI services. . Data cloud company Snowflake is launching the ‘Snowflake Power Platform Connector’ that strengthens interoperability with the MS Power Platform.

As all-round AI technology innovation continues, starting with generative AI, ironically, it seems that a cooperative system for survival and competition has become essential for global companies. In particular, MS, which has responded most quickly to changes caused by AI by early joining hands with Open AI, a leading generative AI company known as ChatGPT, is said to have once again confirmed its leadership in AI technology to the world through this event.

Below is the domestic and foreign cloud-related news that was delivered in Korea last week. If you would like more detailed information on an individual article, you can read the full text by searching the article title.

◆The world’s fastest supercomputer is the U.S. ‘El Capitan’… 13th among Korea’s ‘Top 500’ =‘El Capitan’ at Lawrence Livermore National Laboratory (LLNL) in the United States has become the world’s fastest supercomputer. Oak Ridge National Laboratory, which ranked first in the world’s top 500 supercomputers for three consecutive years announced at the International Supercomputing Conference (SC24) held at the Georgia World Congress Center (GWCC) in Atlanta, USA from the 19th (local time). It surpassed ‘Frontier’. The measured performance of ‘El Capitan’, which ranked first, is 1.742 exaflops (), which is capable of performing 174.2 calculations per second. Among Korea’s supercomputers, Naver’s ‘Sejong’ recorded the highest performance at 40th place. ‘Sejong’ has an actual performance of 32.97 petaflops (PFlops). NHN Cloud’s ‘NHN Cloud Gwangju AI’ is a new entry in this ranking. ‘NHN Cloud Gwangju AI’, which ranked 98th, consists of Dell Technologies PowerEdge XE9680 and NVIDIA H100.

◆The public cloud market is unlocked… “We need to increase domestic competitiveness and strengthen public-private cooperation” =As the government’s policy of easing network separation loosens the bar on the domestic public cloud market, the entry of foreign clouds that have not been able to enter until now is becoming visible. Choi Ji-woong, chairman of the Korea Cloud Industry Association (CEO of KT Cloud), held a press conference at KT Cloud headquarters in Gangnam-gu, Seoul on the 19th and said, “Ultimately, the answer is for domestic cloud service providers (CSPs) to improve quality, but the problem is that they do not have astronomical amounts of money like big techs. “The point is that we don’t have the stamina to pour in all the resources and manpower,” he pointed out. Therefore, Chairman Choi believes that additional support from the government as well as corporate efforts is needed to increase the competitiveness of domestic cloud. In particular, Chairman Choi emphasized that CSPs should not try to provide all services directly, but should create a virtuous cycle that revitalizes the entire cloud ecosystem, including management service providers (MSPs).

◆Hyun Shin-kyun 2nd LG CNS, momentum for IPO and new business next year=LG CNS will continue the leadership of Hyun Shin-kyun next year. On the 21st, CEO Hyun was promoted from vice president to president through the LG Group personnel reshuffle conducted by each affiliate. The current CEO, who was appointed to LG CNS in 2022, was initially focused on reappointment as the company’s listing was imminent. He is evaluated to have been recognized for his management ability and business performance by not only retaining his position but also being promoted to president. Currently, LG CNS is preparing for listing on the stock market (KOSPI) and its over-the-counter market value is approaching 10 trillion won. In the case of IPO, the preliminary review for listing is currently in progress, and LG CNS will submit a securities report as soon as the preliminary review, which usually takes about two months, is passed and accelerate the listing process, including predicting demand for institutional investors, confirming the public offering price, and applying for general investors. It is planned. If the listing is successfully completed, it will be the first LG Group affiliate to be listed in three years since LG Energy Solutions in January 2022.

◆Hancom sends shareholder letter in the second half of the year… “Cloud- and AI-centered global growth” =Hangul and Computer presented a global growth strategy centered on cloud and AI through a shareholder letter in the second half of this year. Hancom said, “This year, we are diversifying our business portfolio by expanding our cloud software-as-a-service (SaaS) business and establishing an AI value chain. In particular, we are expanding our technology to various industrial fields while maximizing scalability by moving away from the existing completed application form with a technology modularization strategy.” “We are expanding our and human assets,” he said. He added, “We are rapidly strengthening our core capabilities by utilizing data-based cloud and AI technologies, and thanks to the remarkable growth of new businesses such as cloud, we are recording the highest ever increases in both sales and operating profit every quarter.” To expand its global business, Hancom is also pursuing technology network cooperation with European AI company Mistral AI.

◆Naver establishes JV with Saudi Arabian National Housing Corporation… Middle East General Corporation =Naver is pursuing the establishment of a joint venture with NHC (National Housing Company), a state-run company under Saudi Arabia’s Ministry of Autonomy, Administration and Housing. A Saudi Arabian regional business unit JV will be established with Naver Arabia (tentative name), which will oversee the Middle East business. The new joint venture will carry out the core business of operating and commercializing the digital twin platform in Saudi Arabia together with NHC under Naver Arabia. It happens. This will be Team Naver’s first business entity to develop a technology platform business in the Middle East. NHC is a state-owned enterprise under the Saudi Ministry of Autonomous Administration and Housing, and 70% of real estate transactions in the country are conducted through NHC. Through this JV, the two sides plan to promote an urban public monitoring platform and a map-based super app for public administration purposes in addition to operating and commercializing Saudi Arabia’s digital twin platform.

* How can businesses ensure responsible ​and ethical implementation of Microsoft’s AI tools, ‍especially considering potential biases and impacts on workforces?

## Open-ended Questions for Discussion on Microsoft’s AI Integration

This article details Microsoft’s recent moves in the AI space, particularly around generative AI and the “AI agent” concept. Here⁤ are⁣ some open-ended questions, ‌divided into thematic sections, to encourage discussion:

**I. AI Adoption &‍ ROI:**

*⁣ Microsoft​ claims​ a 3.7x average ROI from AI adoption. What factors contribute to⁢ this, and do you think these benefits are accessible to all businesses?

* Should companies prioritize‍ AI adoption primarily based on potential ROI, or are there other ethical and ‌societal considerations to weigh?

* How can companies accurately measure‌ the ROI of AI adoption for different departments, like sales or marketing?

**II.⁣ The Future of Work with ⁢”AI Agents”:**

* Microsoft ‌envisions a future ‍ where “AI agents” automate tasks and augment human capabilities. How might this change the nature of work and ⁣the skills required for future jobs?

* What are ‌the potential benefits and drawbacks of relying heavily on AI agents in daily workflows?

* How‌ can we ensure that AI ‍agents are designed and implemented in a ‌way ⁤that benefits both workers and companies?

**III. The “AI Scaling Law” Debate:**

* CEO Satya Nadella ⁤acknowledges the debate about AI scaling limitations. Do you think the pace​ of AI development will eventually hit a plateau,⁤ or will ⁣we find new ways ⁤to overcome these limits?

* What are some potential innovative solutions for pushing⁤ beyond the current limitations⁣ of AI scaling?

*​ How might this debate about AI scaling ​impact future research and investment in‍ AI technology?

**IV. Global ⁢Competition and Collaboration in AI:**

* ​The article highlights partnerships between Microsoft and other companies like Red Hat,‍ TeamViewer, and Snowflake. ⁣How‍ does ⁣this ecosystem of collaboration shape the development and accessibility of AI technology?

* What are the potential risks and benefits of‍ this increasingly interconnected AI landscape?

* In⁤ the race for AI‍ dominance, how do you think competition and cooperation will ultimately shape the development of responsible and ethical AI?

**V. Ethical Implications of Generative AI:**

*⁣ As AI becomes more generative, what are the potential ethical challenges we need‍ to address, such as bias, misinformation, and​ job displacement?

* How can we ensure that generative AI is used responsibly and ethically across different industries ⁢and applications?

* What role should governments and regulatory bodies play in shaping ⁣the development and deployment of generative AI?

These open-ended questions are intended to spark thoughtful discussion and encourage a deeper understanding of the complex ‍implications of Microsoft’s ⁤ambitious AI initiatives.