Chainfeeds Summary:
By December 2026, 30% of international payments will be issued through stablecoins.
Article source:
https://x.com/ttunguz/status/2003249842127057262
Article Author:
Tomasz
Opinion:
Tomasz Tunguz: 1) Enterprises are paying more for AI agents than for human labor. This trend is already happening in the consumer sector. For repetitive tasks in the business, agents will demand a similar premium as enterprises consider recruitment, training, and management costs. 2) 2026 will be a record year for liquidity. SpaceX, OpenAI, Anthropic, Stripe, and Databricks will go public, with SpaceX and OpenAI ranking among the top ten in history. After more than four years of funding shortages, pent-up demand is finally exploding. Fear of the potential disruption caused by rapidly developing AI systems has prompted existing companies to make over $25 billion in defensive acquisitions, preferring to buy rather than develop in-house. 3) Vector databases are experiencing a resurgence as infrastructure in the AI stack. Multimodal models and world/state space models require new data architectures. Vector database revenue is surging, becoming the link between underlying models and enterprise data. 4) AI models will be able to perform tasks continuously for more than a workday. According to METR research, the time it takes for AI to complete tasks doubles every seven months. Current cutting-edge models are already capable of completing tasks in about one hour. Based on this trend, it is projected that by the end of 2026, AI agents will be able to independently execute workflows exceeding eight hours, fundamentally changing how company projects are staffed. 5) AI budgets are under scrutiny for the first time. Purchasing committees and boards will question AI spending. Smaller language models and open-source alternatives will become increasingly popular as research labs specialize them to provide state-of-the-art performance for specific tasks while reducing costs. Developers are increasingly opting for these alternatives because they offer up to 10x cost reductions. 6) Google surpasses competitors with breakthroughs across a broad range of AI areas. No other company has achieved breakthroughs in multiple areas: cutting-edge models, on-device inference, video generation, open-source weights, and search integration. Google has set the pace for technological advancement, forcing OpenAI, Anthropic, and xAI to focus on responding to this challenge. The era of each lab competing in every area is over. [Original text in English]
Content source 
