BTC $63,053.13 -0.04%
ETH $1,880.73 -0.14%
BNB $605.34 -0.86%
XRP $1.00 -0.41%
SOL $75.45 -0.05%
TRX $0.3307 -0.56%
DOGE $0.0697 -0.68%
ADA $0.1768 -2.16%
BCH $203.03 -1.49%
LINK $9.36 -0.91%
HYPE $57.11 +1.37%
AAVE $86.28 -0.80%
SUI $0.6757 -1.36%
XLM $0.1568 -1.47%
ZEC $486.52 -1.36%
BTC $63,053.13 -0.04%
ETH $1,880.73 -0.14%
BNB $605.34 -0.86%
XRP $1.00 -0.41%
SOL $75.45 -0.05%
TRX $0.3307 -0.56%
DOGE $0.0697 -0.68%
ADA $0.1768 -2.16%
BCH $203.03 -1.49%
LINK $9.36 -0.91%
HYPE $57.11 +1.37%
AAVE $86.28 -0.80%
SUI $0.6757 -1.36%
XLM $0.1568 -1.47%
ZEC $486.52 -1.36%

spending

All
Article
Flash

Berkshire's first major portfolio adjustment revealed, spending $17 billion in the second quarter to bet on Google

Berkshire Hathaway submitted its 13F holdings report for the second quarter of 2026 to the U.S. SEC. The data shows that in the second quarter after Warren Buffett stepped down, the company's investment portfolio underwent significant adjustments, with a large purchase of Google’s parent company Alphabet, while reducing positions in the financial and consumer sectors.As of June 30, 2026, the total market value of Berkshire's stock holdings rose to $29.9 billion, up from $26.3 billion in the previous quarter. In the second quarter, the company added one new position, increased holdings in seven stocks, reduced holdings in six stocks, and completely sold out of one target, with the top ten holdings accounting for as much as 88.74%.Among them, Alphabet became the biggest highlight. In the second quarter, Berkshire cumulatively increased its holdings of Alphabet Class A and Class C shares by approximately 48.1 million shares, with the new holdings valued at over $17 billion, pushing Google to replace Bank of America as Berkshire's fourth-largest holding. Currently, its top five holdings are Apple, American Express, Coca-Cola, Alphabet, and Bank of America. In addition to Google, Berkshire also slightly increased its holdings in Delta Air Lines, Lennar, and Macy's. The increase in Delta Air Lines has attracted attention, as the market believes this move may reflect the company's optimism about the recovery of air travel demand and improvement in corporate operations.On the reduction side, Berkshire focused on cutting positions in the financial and consumer sectors in the second quarter. Among them, Bank of America saw a reduction of about 30.2 million shares, with the holding ratio decreasing by 5.89%, corresponding to a market value of about $1.72 billion, making it the largest reduction target; First Capital Financial reduced about 4.2 million shares, with the holding ratio decreasing by about 58%; at the same time, it reduced about 11 million shares of Kroger, with the holding size decreasing by about 22%.The market believes that Berkshire ended a streak of 14 consecutive quarters of net stock selling in the second quarter and net bought nearly $20 billion in stocks, indicating that the new leader Greg Abel is pushing the portfolio towards a technology growth direction, marking a shift in asset allocation in the "post-Buffett era."

Coinbase: Has reduced AI spending by nearly 50% and is trying to default to adopting open weight models

Coinbase CEO Brian Armstrong published an article introducing the company's latest progress in AI cost optimization.Armstrong stated that as the usage of AI and Token consumption continues to grow, the key to controlling costs is not to restrict employee usage or frequently send budget reminders, but to optimize default model selection, task routing mechanisms, and caching strategies.He revealed that Coinbase is trying to use open-weight models such as GLM 5.2 and Kimi 2.7 as default options through an internal LLM gateway, while still allowing engineers to choose other models based on specific task requirements. Data shows that 91% of the company's employees have never reached the AI usage quota limit, so Coinbase has not chosen to tighten quotas but instead improved overall efficiency through lower-cost model solutions.In terms of model routing, Coinbase preprocesses prompts and, combined with cache hit rates and the pricing of different models, automatically assigns tasks to the most suitable model. Armstrong believes that complex tasks such as planning and reasoning may require support from cutting-edge models, but execution tasks do not necessarily need to invoke higher-cost models. In the future, the model selection process should be more automated by AI rather than relying on manual decisions.Additionally, he pointed out that cache hit rate is one of the important factors affecting AI costs. Coinbase has incorporated a cache-aware mechanism into the request process to improve the reuse rate of historical results. For example, in the case of LibreChat, after optimizing the caching solution, its cache hit rate has increased from 5% to 60%.Armstrong also stated that the company requires engineers to keep context as concise as possible, including starting new sessions when switching tasks, narrowing the context scope of files, and closing unused tools, to reduce unnecessary Token consumption.According to him, through these measures, Coinbase has successfully reduced AI spending by nearly 50%, while Token usage continues to grow.
app_icon
ChainCatcher Building the Web3 world with innovations.