Join our daily and weekly newsletters for the latest updates and exclusive content in the industry’s leading AI coverage. Learn more
Stanford Human-Center Artificial Intelligence Institute (Two) The global development of the EU has issued a 2025 AI index report that provides information managing analysis. Hai has developed a report on the EI over the past few years the first criterion Comes in 2022. There is no need to say, a lot has changed.
2025 sheet is loaded with statistics. Between the best finds:
- The United States, in 2024, 40 notable AI models, China (15) and Europe (3) were much ahead.
- Training calculation for AI models is about five months in almost every five months, database every eight months.
- The cost of the results of the AI fell sharply – 280 times decrease from 2022 to 2024.
- In 2024, the global private AI investment reached $ 252.3 billion, a 26% increase.
- 78% of organizations using AI (above 55% in 2023).
The company offers critical concepts by changing the scenery of heads, reports, model performance, investment trends, application difficulties and technology.
Here are five basic takeay roads for the heads of the AI index.
1. The democratization of the AI power is accelerated
Perhaps the most surprising finding was a high quality AI more suitable and accessible. Once the expense barrier that restricts an advanced AI to technical giants. How different is the find 2024 Stanford sheet found.
« I did not look at how cheaper, cheaper, cheaper and the future in the future, » Nestor Maslej said in the Hai Research Manager Venturebeat for the AI Index. « Although training costs remain high, now we see a world despite the reduction of border models of high quality development costs. »
The report sharply increases the landslide: The cost of the end of November 2022 in November 2022 in 2022 in 2022 in 2024, more than 280 times decreased by $ 0.07.
The performance convergence between closed and open weight models is equally important. The gap between the best closed models (such as GPT-4) and leading open models (such as LLAMA) in 2025, 2025 in 2025, in 2025, less than 8.0%.
IT leader action product: Revalate your AI purchase strategy. Earlier, organizations spoke about advanced AI opportunities now have access to open weight models or cheaper trading APIs.
2. The space between the AI adoption and value realization remains significant
The report now has 78% of organizations in at least one business function from the EU (above 55% in 2023), a real business effect on the adoption.
Asked about the meaningful ROI in the measurement, while Maslej acknowledged: « We are intending to explore more, there are limited information about organizations that earned. This is the critical area of the analysis that intends to explore later. »
The report shows that most organizations of the huge financial development of the EU are most organizations. For example, 47% of enterprises using a generative EU in the strategy and corporate financial report are growing, but usually in lower levels.
IT leader action product: Focus on measurable use, which is a clear ROI potential than extensive application. Consider developing stronger AI management and measurement frames to better watch the value creation.
3. Special work functions indicate stronger financial income than AI
The report features granular concepts, if the work functions are the most important financial effects on the application of the AI.
« At the price side, the AI, supply chain and service operations functions are most benefited, » Maslej said. « See the biggest earnings of the income side, strategy, corporate finance and supply chain functions. »
In particular, 61% of organizations using generative AI in the supply chain and inventory management report are growing in the strategy and corporate financial report. Service operations and marketing / sales also show strong potential to create value.
IT leader action product: Prioritize the EU investments in the report, which shows the most important financial income in the report. Supply chain optimization, service operations and strategic planning occur as high potential areas for initial or expanded AI placement.
4. AI shows strong potential to equalize the operation of labor force
One of the most interesting results affects the levels of affecting the AI’s workforce productivity. Multiple work in the report, AI tools are disproportionate to low-qualified workers.
In the context of customer support, lower skills workers, in the event of minimal development of highly skilled workers, 34% have experienced 34% productivity. Similar samples appeared in consultations (43% 16.5% earnings) and the software (21-40% of 7-46%).
« In general, these studies show that the EU has a strong positive impact on productivity and highly qualified workers than higher skilled workers, » Maslej said.
IT leader action product: Review the placement of AI as a workforce development strategy. AI Assistants can help the play area among small and large workers, which improve the performance of the general team, potentially touching skills.
5. The application of the AI responsible remains as a reality, not a reality
Despite the growing awareness of AI risks, the report reveals an important space between recognition and reduction of risks. 66% of organizations consider a risk related to the EU in terms of cyberem, only 55% actively reduce it. Similar gaps for adjustment compatibility (63% 38%) and the violation of intellectual property (57% 38%).
These findings increase by 56.4% in a record record in 2024 in 233, which increased by 56.4% to the background of increased organizations. Organizations are facing real results because they are uninhabitable EU experience.
IT leader action product: Do not delay healthy responsible AI management. When technical opportunities progress quickly, the report offers the lack of effective risk reduction strategies of most organizations. It can be a competitive advantage, not the load of compliance, not loading the frames.
Looking forward
Stanford AI Index report is more accessible and skillful to grow rapidly and competent, organizations still fight to capitalize their potential.
Strategic imperative for IT leaders: Notice the applications applied to the target with the target ROI, emphasize responsible management and leverage AI to increase the capacity of labor.
« In this turn, it points to more accessible and I believe that the wide wave of the EI can be on the horizon, » Maslej said.
Source link
Leave a Reply