Critics of artificial intelligence (AI) contend that the current AI boom is overstated, arguing that much of the recent advancements are the result of exploiting easily achievable milestones—often referred to as "low-hanging fruit." They suggest that future progress will be significantly slower, more challenging, and costly, implying that the most significant breakthroughs may have already occurred.
Contrary to this perspective, the latest findings from Stanford University’s Institute for Human-Centered Artificial Intelligence present a different narrative. According to the 2026 AI Index Report, AI capabilities are not stagnating; instead, they are advancing rapidly and reaching an ever-growing audience. The report highlights measurable improvements across various benchmarks over time, indicating that many AI systems are achieving human-level performance.
Notably, advancements in software engineering demonstrate that AI systems have progressed from solving approximately 60% of related problems to nearly complete reliability. Additionally, the benchmark "Humanity’s Last Exam," which assesses expert-level reasoning, saw an improvement in success rates from 8.8% to over 38% in just a year, with some systems surpassing the 50% mark. Cybersecurity accuracy rates also soared from 15% to 93% during the same period.
The report also indicates a significant increase in global AI computational capacity and investments, which reached nearly $582 billion in 2025. The rapid adoption of generative AI is reportedly outpacing the early years of personal computers and the internet, with 88% of organizations reportedly integrating AI technologies.
While the report does acknowledge existing limitations, such as advanced systems struggling with elementary tasks, it emphasizes the long-term trajectory of AI technology as a foundational element of modern infrastructure.
Why this story matters: The findings challenge prevailing skepticism about the future of AI, highlighting significant ongoing advancements.
Key takeaway: AI capabilities are accelerating, with a notable rise in performance across various domains.
Opposing viewpoint: Critics argue that progress is overstated and future advancements will be slower and more costly.
