White House advisor on cryptocurrencies and artificial intelligence David Sachs believes that concerns about mass layoffs due to the development of AI are exaggerated. In his opinion, even the most advanced AI models still require human oversight and expert involvement to deliver business value.
This is reported by Finway
Artificial Intelligence as a Tool, Not a Complete Replacement for Specialists
On the social network X (formerly Twitter), David Sachs explained that artificial intelligence serves as a “middle link” in work processes. Leadership, crucial decision-making, and oversight remain solely in human hands. Sachs emphasized that processes involving AI still require an iterative approach, verification, and adjustments by specialists to achieve real effectiveness.
“AI models still need to be fine-tuned and verified, often repeatedly, to achieve business outcomes. As Balaji Srinivasan notes, AI is a middle of the middle, not a full cycle. Humans remain at key stages of the process, while artificial intelligence takes on intermediate tasks.”
Sachs’s views have become particularly relevant following the publication of a Microsoft Research study. It identified professions most vulnerable to automation through AI. These fields include journalism, analytics, and technical writing—professions that are also actively utilized in the cryptocurrency industry. According to an analysis of over 200,000 Bing Copilot chats, AI is most frequently used for gathering information, generating texts, and providing consultations.
The Impact of AI on the Crypto Industry and Labor Market
According to Microsoft estimates, professions related to reporting and copywriting have the highest level of automation, with AI applicability ratings of 0.38–0.39. Market analysts and data specialists received ratings of 0.35–0.36. This indicates an increased risk of automation specifically in these areas of activity.
The situation is also relevant for the cryptocurrency sector. In July 2025, experts noted a significant reduction in the number of open vacancies in the industry. This trend coincided with official data from the U.S. Department of Labor, which reported that only 73,000 new jobs were created in the month, instead of the expected 100,000 according to Dow Jones forecasts.
David Sachs also highlighted the perspective of former Coinbase CTO Balaji Srinivasan. He emphasized that the current generation of AI is not fully autonomous and rather enhances human productivity than completely replaces it.
“Today’s AI is not truly agentic. It enables more tasks to be performed, but it does not operate fully independently,” Srinivasan stated.
According to Srinivasan, the main competition occurs between different AI models. For instance, the Midjourney system has displaced Stable Diffusion in the field of image generation, while GPT-4 has replaced GPT-3 among general-purpose language models. This supports Sachs’s conclusion that the effectiveness of artificial intelligence application directly depends on human oversight and skilled management.
It was previously reported that Google introduced a new advanced AI model—Gemini Deep Think, designed for more complex logical tasks and reasoning.