Oct 2026· Advances in Economics Management and Political Sciences
Business Strategies and Innovation
Abstract
With the rapid development of artificial intelligence, large-scale technology firms generally face the challenge of balancing the stability of mature businesses with new growth opportunities. Taking Microsoft as an example, this paper discusses the relationship between business portfolio diversification and AI-driven growth strategy. This article focuses on how Microsoft can maintain revenue stability through its business segments—such as productivity software, cloud computing, personal computing, and gaming—while also creating new growth opportunities through AI-related businesses. This paper uses a case study approach, along with Microsoft annual reports, investor information, and public market data, to analyze business-segment revenue, AI-related growth, stock returns, volatility and market risk characteristics. This article does not regard AI as an independent product, but analyzes how it is embedded in Microsoft's overall business portfolio through Azure, Microsoft 365, GitHub and other channels. The results show that Microsoft's diversified business segments provide a relatively stable revenue base for AI‑driven growth, while cloud computing and AI-related businesses further enhance the company's growth potential. From 2021 to 2025, Microsoft's cumulative return and annualized return were 131.53% and 18.33%, respectively, higher than the S&P 500's 84.98% and 13.12%; however, Microsoft's annualized volatility was also higher. The research results show that business diversification can support Microsoft's AI-driven growth, but this higher growth potential is accompanied by more pronounced market risks. Therefore, Microsoft's future portfolio optimization needs to maintain a balance between AI investment, profitability, cash flow stability and risk exposure.
This paper investigates developing real-world perception for active agents, proposes Gibson Environment for this purpose, and showcases a set of perceptual tasks learned therein.
F. Xia, Amir Zamir, Zhi-Yang He et al.· 2018 IEEE/CVF Conference on...· 1,069 citations· ⚡112
GAOKAO-Bench is introduced, an intuitive benchmark that employs questions from the Chinese GAOKAO examination as test samples, including both subjective and objective questions that contribute a robust evaluation benchmark for future large language models and offers valuable insights into the advantages and limitations...
Xiaotian Zhang, Chun-yan Li, Yi Zong et al.· arXiv.org· 216 citations· ⚡17
An algorithm is proposed which leverages disentangled semantic factors to generate adversarial perturbation by altering controlled semantic attributes to fool the learner towards various "adversarial" targets.
Haonan Qiu, Chaowei Xiao, Lei Yang et al.· arXiv.org· 208 citations· ⚡17
This work proposes a HALO that directly maximizes the utility of generations instead of maximizing the log-likelihood of preferences, and suggests that the best loss depends on the inductive biases most appropriate for a given setting, an oft-overlooked consideration.
This work investigates the possibilities of using LLMs in a resume screening setting via a document retrieval framework that simulates job candidate selection and finds that the MTEs are biased, significantly favoring White-associated names in 85% of cases and female-associated names in only 11.1% of cases.
This work shows that orders of magnitude enhancement in performance could be obtained by a combination of hardware improvements and tight quantum-HPC integration and introduces high-performance architectures for quantum-probabilistic computing with custom-designed accelerators to tackle today's industry-scale classical...
Masoud Mohseni, Artur Scherer, K. Johnson et al.· arXiv.org· 121 citations· ⚡9