Trump might claim China tariff victory – but this is Capitulation Day
Stocks were up and bond yields were higher after the US treasury secretary Scott Bessent's early morning press conference in Geneva, where he has been holding talks with China.
As with the UK 'trade deal' last week, the US is not reverting to the status quo before Trump arrived in the White House.
Related: China and US agree 90-day pause to trade war initiated by Donald Trump
Instead, tariffs on Chinese goods will be cut from 145% to 30% – initially for a 90-day period. In return, China has cut its own tariffs on US imports to 10%, from the 125% it had imposed in retaliation against the White House.
That still marks a big shift in the terms of trade between the two countries since before Trump came to power, but falls far short of what was in effect a trade embargo.
The two sides have pledged to keep talking, but there was no reference in the statement put out by the White House to other gripes it has previously raised about China, including the weakness of the yuan.
Instead, the statement hailed 'the importance of a sustainable, long-term and mutually beneficial economic and trade relationship'. The language was rather different to Trump's Liberation Day speech, about the US being 'looted, pillaged, raped and plundered by nations near and far'.
In other words, the president has caved. He may have been swayed by market wobbles but it seems more plausible that dire warnings from retailers about empty shelves – backed up by data showing shipments into US ports collapsing – may have strengthened the hands of trade moderates in the administration.
Confronted with warnings of a shortage of toys, Trump told reporters that children should be happy with 'two dolls instead of 30 dolls', and they might 'cost a couple bucks more' than usual. But it is difficult to imagine even this most bullish of presidents withstanding the attacks that would come his way if he began to be seen as responsible for Covid-style shortages of key goods in the world's largest economy.
Instead, the White House seems to have opted for tactical retreat. The China-US conflict was always the hottest theatre of confrontation in Trump's trade war, with a longer history and deeper public support than his quixotic attacks on Mexico and Canada.
If Trump is indeed ready to give in even with Beijing, it sends a signal that some of the other aggressive aspects of his trade policy may be negotiable.
What Bessent and his Chinese counterparts have not erased, however, is the corrosive uncertainty that has gripped investors across the global economy since Trump's 'Liberation Day' tariff announcement.
China tariffs have only been slashed temporarily, for now and many other countries are still awaiting negotiations on where their tariff levels will end up, after that other 90-day pause, on Trump's 'reciprocal' levies, due to end in July.
Meanwhile, companies throughout the global trading system are left wondering which particular iteration of the policy is likely to stick, and may well be tempted to continue working around the US, where possible.
And with 30% tariffs remaining on Chinese exports to the US, the bigger picture remains of two great economic powers pulling apart.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
TechCrunch
a minute ago
- TechCrunch
OpenAI launches two ‘open' AI reasoning models
OpenAI announced Tuesday the launch of two open-weight AI reasoning models with similar capabilities to its o-series. Both are freely available to download from the online developer platform, Hugging Face, the company said, describing the models as 'state-of-the-art' when measured across several benchmarks for comparing open models. The models come in two sizes: a larger and more capable gpt-oss-120b model that can run on a single Nvidia GPU, and a lighter-weight gpt-oss-20b model that can run on a consumer laptop with 16GB of memory. The launch marks OpenAI's first 'open' language model since GPT-2, which was released more than five years ago. In a briefing, OpenAI said its open models will be capable of sending complex queries to AI models in the cloud, as TechCrunch previously reported. That means if OpenAI's open model is not capable of a certain task, such as processing an image, developers can connect the open model to one of the company's more capable closed models. While OpenAI open-sourced AI models in its early days, the company has generally favored a proprietary, closed-source development approach. The latter strategy has helped OpenAI build a large business selling access to its AI models via an API to enterprises and developers. However, CEO Sam Altman said in January he believes OpenAI has been 'on the wrong side of history' when it comes to open sourcing its technologies. The company today faces growing pressure from Chinese AI labs — including DeepSeek, Alibaba's Qwen, and Moonshot AI —which have developed several of the world's most capable and popular open models. (While Meta previously dominated the open AI space, the company's Llama AI models have fallen behind in the last year.) In July, the Trump Administration also urged U.S. AI developers to open source more technology to promote global adoption of AI aligned with American values. Techcrunch event Tech and VC heavyweights join the Disrupt 2025 agenda Netflix, ElevenLabs, Wayve, Sequoia Capital — just a few of the heavy hitters joining the Disrupt 2025 agenda. They're here to deliver the insights that fuel startup growth and sharpen your edge. Don't miss the 20th anniversary of TechCrunch Disrupt, and a chance to learn from the top voices in tech — grab your ticket now and save up to $675 before prices rise on August 7. Tech and VC heavyweights join the Disrupt 2025 agenda Netflix, ElevenLabs, Wayve, Sequoia Capital — just a few of the heavy hitters joining the Disrupt 2025 agenda. They're here to deliver the insights that fuel startup growth and sharpen your edge. Don't miss the 20th anniversary of TechCrunch Disrupt, and a chance to learn from the top voices in tech — grab your ticket now and save up to $675 before prices rise. San Francisco | REGISTER NOW With the release of gpt-oss, OpenAI hopes to curry favor with developers and the Trump Administration alike, both of which have watched the Chinese AI labs rise to prominence in the open source space. 'Going back to when we started in 2015, OpenAI's mission is to ensure AGI that benefits all of humanity,' said OpenAI CEO Sam Altman in a statement shared with TechCrunch. 'To that end, we are excited for the world to be building on an open AI stack created in the United States, based on democratic values, available for free to all and for wide benefit.' (Photo by) Image Credits:Tomohiro Ohsumi / Getty Images How the models performed OpenAI aimed to make its open model a leader among other open-weight AI models, and the company claims to have done just that. On Codeforces (with tools), a competitive coding test, gpt-oss-120b and gpt-oss-20b score 2622 and 2516, respectively, outperformed DeepSeek's R1 while underperforming o3 and o4-mini. OpenAI's open model performance on codeforces (credit: OpenAI). On Humanity's Last Exam, a challenging test of crowd-sourced questions across a variety of subjects (with tools), gpt-oss-120b and gpt-oss-20b score 19% and 17.3%, respectively. Similarly, this underperforms o3 but outperforms leading open models from DeepSeek and Qwen. OpenAI's open model performance on HLE (credit: OpenAI). Notably, OpenAI's open models hallucinate significantly more than its latest AI reasoning models, o3 and o4-mini. Hallucinations have been getting more severe in OpenAI's latest AI reasoning models, and the company previously said it doesn't quite understand why. In a white paper, OpenAI says this is 'expected, as smaller models have less world knowledge than larger frontier models and tend to hallucinate more.' OpenAI found that gpt-oss-120b and gpt-oss-20b hallucinated in response to 49% and 53% of questions on PersonQA, the company's in-house benchmark for measuring the accuracy of a model's knowledge about people. That's more than triple the hallucination rate of OpenAI's o1 model, which scored 16%, and higher than its o4-mini model, which scored 36%. Training the new models OpenAI says its open models were trained with similar processes to its proprietary models. The company says each open model leverages mixture-of-experts (MoE) to tap fewer parameters for any given question, making it run more efficiently. For gpt-oss-120b, which has 117 billion total parameters, OpenAI says the model only activates 5.1 billion parameters per token. The company also says its open model was trained using high-compute reinforcement learning (RL) — a post-training process to teach AI models right from wrong in simulated environments using large clusters of Nvidia GPUs. This was also used to train OpenAI's o-series of models, and the open models have a similar chain-of-thought process in which they take additional time and computational resources to work through their answers. As a result of the post-training process, OpenAI says its open AI models excel at powering AI agents, and are capable of calling tools such as web search or Python code execution as part of its chain-of-thought process. However, OpenAI says its open models are text-only, meaning they will not be able to process or generate images and audio like the company's other models. OpenAI is releasing gpt-oss-120b and gpt-oss-20b under the Apache 2.0 license, which is generally considered one of the most permissive. This license will allow enterprises to monetize OpenAI's open models without having to pay or obtain permission from the company. However, unlike fully open source offerings from AI labs like AI2, OpenAI says it will not be releasing the training data used to create its open models. This decision is not surprising given that several active lawsuits against AI model providers, including OpenAI, have alleged that these companies inappropriately trained their AI models on copyrighted works. OpenAI delayed the release of its open models several times in recent months, partially to address safety concerns. Beyond the company's typical safety policies, OpenAI says in a white paper that it also investigated whether bad actors could fine-tune its gpt-oss models to be more helpful in cyber attacks or the creation of biological or chemical weapons. After testing from OpenAI and third-party evaluators, the company says gpt-oss may marginally increase biological capabilities. However, it did not find evidence that these open models could reach its 'high capability' threshold for danger in these domains, even after fine-tuning. While OpenAI's model appears to be state-of-the-art among open models, developers are eagerly awaiting the release of DeepSeek R2, its next AI reasoning model, as well as a new open model from Meta's new superintelligence lab.
Associated Press
a minute ago
- Associated Press
The Trump administration dismisses most on a federal board overseeing Puerto Rico's finances
SAN JUAN, Puerto Rico (AP) — The Trump administration has dismissed five out of seven members on Puerto Rico's federal control board that oversees the U.S. territory's finances, sparking concern about the future of the island's fragile economy. The five fired are all Democrats. A White House official told The Associated Press on Tuesday that the board 'has been run inefficiently and ineffectively by its governing members for far too long and it's time to restore common sense leadership.' Those fired are board chairman Arthur Gonzalez, along with Cameron McKenzie, Betty Rosa, Juan Sabater and Luis Ubiñas. The board's two remaining members — Andrew G. Biggs and John E. Nixon — are Republicans. Sylvette Santiago, a spokesperson for the board, said none of those fired had received notifications ahead of their dismissal. The board was created in 2016 under the Obama administration, a year after Puerto Rico's government declared it was unable to pay its more than $70 billion public debt load and later filed for the biggest municipal bankruptcy in U.S. history. In remarks to the AP, the White House official claimed the board had operated ineffectively and in secret and said it 'shelled out huge sums to law, consulting and lobbying firms.' The official, who spoke on condition of anonymity to discuss the subject, also accused the board's staff of receiving 'exorbitant salaries.' The board spokesperson did not return a message seeking additional comment. Puerto Rico is struggling to restructure more than $9 billion in debt held by the state's Electric Power Authority, with officials holding bitter mediations with creditors demanding full payment. It's the only Puerto Rico government debt pending a restructuring, with the White House official accusing the board of preferring to 'extend the bankruptcy.' In February, the board's executive director, Robert Mujica Jr., said it was 'impossible' for Puerto Rico to pay the $8.5 billion that bondholders are demanding. He instead unveiled a new fiscal plan that proposed a $2.6 billion payment for creditors. The plan does not call for any rate increases for an island that has one of the highest power bills in any U.S. jurisdiction as chronic power outages persist, given the grid's weak infrastructure. Alvin Velázquez, a bankruptcy law professor at Indiana University, said he worries the dismissal of the board members could spark another crisis in Puerto Rico. 'This is really about getting a deal out of (the power company) that is not sustainable for the rate payers of Puerto Rico,' he said. Velázquez, former chair for the unsecured creditors committee during the bankruptcy proceedings, also questioned if the dismissals are legal, since board members can only be removed for just cause. 'What's the cause?' he said. 'What you're going to see is another instance in which the Trump administration is taking on and testing the courts.' The dismissals were first reported by the Breitbart News Network, a conservative news site.
The Hill
a minute ago
- The Hill
OpenAI, Google, Anthropic AI models added to government purchasing system
Artificial intelligence (AI) models from OpenAI, Google and Anthropic have been added to a government purchasing system, allowing federal agencies to buy and use the AI products. The General Services Administration (GSA) announced Tuesday that ChatGPT, Gemini and Claude had been added to the agency's Multiple Award Schedule for purchase. 'America's global leadership in AI is paramount, and the Trump Administration is committed to advancing it,' GSA acting administrator Michael Rigas said in a statement. 'By making these cutting-edge AI solutions available to federal agencies, we're leveraging the private sector's innovation to transform every facet of government operations,' he continued. This follows the addition of xAI's Grok to the GSA schedule, which it announced last month after unveiling a new suite of products for U.S. government customers and scoring a Pentagon contract alongside the three other tech firms. The agency pointed to President Trump's AI Action Plan for the new additions to its purchasing system. The AI framework, released last month, called for accelerating AI adoption in the federal government. It specifically advocated for the creation of an AI procurement toolbox managed by the GSA that would 'allow any Federal agency to easily choose among multiple models in a manner compliant with relevant privacy, data governance, and transparency laws.' The recommendations for federal AI adoption represent one small portion of Trump's wide-ranging AI Action Plan, which also called for limiting state and federal regulations, fast-tracking permitting for data center and energy construction and creating export packages of U.S. technology.
