Scan to download
BTC $59,444.12 -0.91%
ETH $1,590.50 +0.74%
BNB $551.67 -0.22%
XRP $1.04 -0.21%
SOL $73.98 +2.23%
TRX $0.3191 -1.26%
DOGE $0.0722 -0.82%
ADA $0.1448 -0.01%
BCH $199.84 +1.57%
LINK $7.29 -0.32%
HYPE $65.51 +4.68%
AAVE $90.02 -1.71%
SUI $0.6951 +0.56%
XLM $0.1814 +4.74%
ZEC $398.97 +4.39%
BTC $59,444.12 -0.91%
ETH $1,590.50 +0.74%
BNB $551.67 -0.22%
XRP $1.04 -0.21%
SOL $73.98 +2.23%
TRX $0.3191 -1.26%
DOGE $0.0722 -0.82%
ADA $0.1448 -0.01%
BCH $199.84 +1.57%
LINK $7.29 -0.32%
HYPE $65.51 +4.68%
AAVE $90.02 -1.71%
SUI $0.6951 +0.56%
XLM $0.1814 +4.74%
ZEC $398.97 +4.39%

edu

All
Article
Flash

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.

first_img Japan's large corporate pension funds plan to allocate about 1% to cryptocurrencies and reduce their exposure to the yen

According to CoinPost, Japan's national corporate pension fund plans to start investing in cryptocurrencies in the fiscal year 2026, with an allocation ratio of about 1% of its total operating assets (approximately 21.3 billion yen).The report states that the asset allocation ratio for the fiscal year 2025 is: 80% in yen, 15% in US dollars, and 5% in other currencies. However, in the fiscal year 2026, the yen allocation ratio will decrease to 70%, and a new 10% allocation will be made for currencies from developed countries. The remaining 5% will consist of emerging market currencies, gold, and cryptocurrencies.The main purpose is to diversify currency risk. The fund's executive director, Ai Yuki, stated that due to the potential weakening of the US dollar as a benchmark currency, they decided not to increase their holdings in US dollars and instead use cryptocurrencies like Bitcoin as a hedge against currency depreciation, as Bitcoin has a lower correlation with the US dollar index.After approximately six years of investigation, the fund has determined that the cryptocurrency market has matured as the investor base has expanded. In the future, the fund will continue to explore the possibility of expanding cryptocurrency investments, including funds for arbitrage trading of various cryptocurrencies.

The tech industry is experiencing a wave of AI-driven layoffs, with giants like Oracle and Amazon significantly reducing positions

According to the latest industry reports and corporate disclosure documents, the technology industry is experiencing a large-scale wave of layoffs driven by artificial intelligence (AI) in 2026. Despite several companies achieving record high revenues, major tech giants are intensively restructuring their organizational frameworks to reallocate funds towards AI infrastructure development and to enhance operational efficiency through AI. Data shows that in May of this year, the number of layoffs in the tech industry reached the highest monthly record in years, with AI being the core reason for the layoffs.On the execution level, several leading companies have implemented large-scale personnel reductions. Oracle's latest documents reveal that in the past 12 months, 21,000 employees (approximately 13% of the total workforce) have been laid off due to internal AI technology deployment. Amazon cut 16,000 corporate positions in January this year, with management expecting that the widespread application of generative AI will significantly reduce the demand for traditional roles. While Meta laid off about 8,000 employees, nearly 7,000 were reorganized into core AI business positions. Block significantly reduced its workforce by 4,000, nearly halving its total number of employees to adapt to the flattened operational model brought about by AI tools.Additionally, companies including Cisco (4,000 people), Intuit (3,000 people), Atlassian (1,600 people), Cloudflare (1,100 people), Snap (1,000 people), as well as Coinbase, Salesforce, and others have announced substantial layoff plans related to AI transformation this year. At the same time, although Google, Microsoft, and IBM have not disclosed specific total layoff numbers, they are also continuously advancing rolling job replacements and restructuring linked to AI strategies.

Strategy MSTR's convertible bonds have been reduced from $8.2 billion to $6.7 billion, and Coinbase has become one of the three major cryptocurrency concept stocks ahead of this week's FOMC

According to BBX data, ahead of the FOMC meeting yesterday and under the dual catalyst of the US-Iran agreement, the sentiment for cryptocurrency concept stocks has significantly warmed up. The core dynamics are as follows:Strategy, Inc. (NASDAQ: $MSTR) rose 3.18% to $123.97 on June 15, marking one of several days of recovery; Bitcoin rose to about $64,000 during the same period, but there remains about a 15% discount compared to the company's average price of $75,680 for 843,738 BTC. The company's most important balance sheet action recently came from the SEC 8-K on May 25: repurchasing $1.5 billion in face value convertible bonds maturing in 2029 for about $1.38 billion in cash (completed at about an 8% discount to face value), reducing the convertible bond stock from $8.2 billion to $6.7 billion, generating a BTC Yield of 0.7% and a BTC Gain of about 4,391 BTC; as of May 25, USD reserves were $871 million, and the company stated it would "supplement reserves over time based on market conditions"; since the beginning of 2026, the cumulative BTC Yield is 13.3%. Phong Le (CEO): This transaction "reflects the discipline of using comprehensive capital management tools in debt management"; Saylor previously stated that Strategy still had assets covering all debts when BTC fell to $8,000, implying resilience in extreme scenarios.Coinbase Global, Inc. (NASDAQ: $COIN) was listed by CoinGape in a June 15 research report as one of the "three cryptocurrency concept stocks to watch" ahead of this week's FOMC (the other two being $MSTR and $BMNR); currently, BTC is about $64,000 and ETH is about $1,660, with the market pricing a 97.4% probability of no interest rate hike at the FOMC on June 17 (2.6% for a rate cut, 0% for a rate hike); the significant US-Iran agreement (over the weekend of June 14-15) has driven a comprehensive rebound in risk assets, with a sharp drop in oil prices easing inflationary pressures, providing additional support for the recovery of sentiment in the cryptocurrency market—if the FOMC dot plot does not show unexpectedly hawkish signals, Coinbase's prediction market and institutional custody business are expected to benefit from the dual improvement in trading volume and asset scale brought about by the recovery of BTC trends.
app_icon
ChainCatcher Building the Web3 world with innovations.