Connect with us

America

Silicon Valley startups are turning to Chinese open-source AI models

Published

on

Misha Laskin, a theoretical physicist and machine learning engineer who contributed to the development of some of Google’s most powerful AI models, encountered a concerning picture when examining the American AI landscape earlier this year.

Laskin observed a growing interest among US AI companies in free, customizable, and increasingly powerful open-source AI models.

The vast majority of these models are produced in China and are rapidly gaining ground against their US competitors.

Assessing the current situation, Laskin stated, “These models are not far behind the frontier (the cutting edge of technology). In fact, they are surprisingly close to the frontier. What is coming now is noticeably close to the frontier.”

Following this development, Laskin founded a startup called Reflection AI to offer an open-source American alternative to the Chinese models gaining traction in Silicon Valley.

The founder of the company, which recently reached an $8 billion valuation, said, “You are starting to see signs that open model companies in China are actually pushing the frontier of intelligence and the limits of intelligence technology in general.”

Over the past year, a significant portion of America’s most popular AI startups have turned to Chinese open AI models, which compete with and sometimes replace expensive US systems as the foundation for American AI products.

More than 15 AI startup founders, engineers, and industry experts who spoke to NBC News stated that American companies’ models still hold the lead in terms of capability.

However, experts emphasized that many Chinese systems are cheaper to access, more customizable, and have become sufficiently competent for many use cases over the past year.

Cost and speed advantages are changing preferences

Investors have poured tens of billions of dollars into OpenAI and Anthropic with the expectation that leading American AI companies will dominate the global market.

But the increasing use of free Chinese models by American companies raises questions about how exceptional these models are and whether America’s insistence on a “closed model” approach is flawed.

Michael Fine, head of machine learning at the search company Exa, which is valued at $700 million and backed by established Silicon Valley investors Lightspeed Venture Partners and Nvidia, said that running Chinese models on their own hardware is, in many cases, much faster and cheaper than using large models like OpenAI’s GPT-5 or Google’s Gemini.

Fine described the process:

“We often launch a feature with a closed model, but then we realize it’s too expensive or too slow, and we ask, ‘What tricks do we have up our sleeve to make this faster and cheaper?'”

Fine stated that the solution is often to replace the closed model with an equivalent open model and then run it on their own infrastructure.

Chinese-origin systems like DeepSeek’s R1 and Alibaba’s Qwen models can be used for free because they are “open-source” or “open-weight,” meaning anyone can download, copy, modify, and run them.

These systems differ from “closed” systems accessed through data centers controlled by major tech giants, such as Anthropic’s Claude or OpenAI’s GPT models.

The technology gap is closing fast

For years, the closed-source models from OpenAI and Anthropic performed far better than both American and Chinese open alternatives.

Even open-source initiatives like BloombergGPT, trained by institutions with resources like Bloomberg on their own financial data, lagged behind OpenAI’s closed models in financial knowledge.

However, over the past year, Chinese companies like DeepSeek and Alibaba have made significant technological strides. According to metrics tracked by Artificial Analysis, an independent AI benchmarking company, their open-source products now approach or match the performance of leading closed American models in many areas.

“The gap is really narrowing,” said Lin Qiao, co-creator of PyTorch, the dominant framework for training AI models, and CEO of Fireworks AI, regarding the capability difference between American closed-source and Chinese open-source models.

As a result of this performance increase, platforms like OpenRouter, which allow users to choose between different models, are seeing a shift toward Chinese open-source models.

Jerry Liu, founder of the productivity app Dayflow, estimates that about 40% of his users now prefer to use open-source models.

Dayflow offers an application built on basic tasks like scanning screenshots and summarizing user activity.

Users can choose between Google’s Gemini model and smaller open-source options like Alibaba’s Qwen.

Liu noted that for tasks like describing a user’s screen, the Qwen model is extremely consistent, stating, “Qwen is as good as GPT-5 for my use case.”

Unlike GPT-5 or Gemini, a smaller version of Qwen can be run at a relatively low cost or for free.

Liu mentioned that paying for closed model usage could cost Dayflow up to $1000 per person, making cheaper open-source models critical for the application’s sustainability.

Privacy sensitivity encourages local processing

The open-source models used by Dayflow perform all processing on each user’s own computer. Liu stated that this is attractive to users who do not want to send their data to the cloud for privacy reasons.

Emphasizing his preference for using open-source models on his own device, Liu said, “Would I use a product where my entire screen is beamed to some random guy’s cloud? Never.”

In addition to increased performance, stronger privacy, and lower costs, open-source models are also gaining ground due to ecosystem advantages.

The rising adoption rate among developers and the open-source systems they create encourage more software engineers to use these models.

Antonio Vespoli, co-founder of the browser assistant startup Circlemind AI, said that Chinese models now dominate online developer resources.

There is a practical reason for this: Chinese models like Qwen, which Airbnb CEO Brian Chesky stated they rely on “heavily,” have abundant training guides and community support.

Charles Zedlewski, chief product officer at the AI infrastructure company Together AI, noted that developers now find it simpler and more efficient to start with open models and adapt them with their own data.

Zedlewski stated that companies understand their needs more clearly as they launch their first AI applications.

Of the top 20 models among users of Kilo Code, a popular application that helps software engineers write code, seven are of Chinese origin, and six of them are open-source.

Beijing’s strategic support and production speed

While most of America’s AI developments occur in the private sector and with a closed-model approach led by industry giants like OpenAI and Anthropic, the Chinese government plays a more active role in charting the country’s AI vision.

In a speech on November 1, Chinese President Xi Jinping called for “more cooperation in open-source technologies.”

In March, China’s top economic planning authority announced its intention to support an ecosystem of open-source models.

While Chinese labs generally release their models openly, American companies like OpenAI achieved early success with closed models and have remained committed to that approach.

Furthermore, many Chinese companies are releasing their products at a faster pace than their American competitors.

Alibaba has released a new model roughly every 20 days this year, while the average time between Anthropic’s releases has been 47 days.

Nathan Lambert, a senior research scientist at the Allen Institute for AI and an expert on the open model ecosystem, told NBC News that the recent progress of Chinese models is no coincidence.

“The Chinese are real innovators in AI,” Lambert said.

Lambert, who writes extensively about China’s AI developments on the Substack platform and is considered an expert on China’s open-source ecosystem, added that the balance of power has shifted rapidly in the last 12 months.

Some in Silicon Valley note that American models still hold a significant advantage at the cutting edge of AI capabilities and that closed American models offer a user-friendliness that cumbersome open models cannot match.

Tim Tully, a partner at Menlo Ventures, argued that closed models are still much more capable and generally more useful:

“The tools are better, the productivity is better, the agent frameworks being built and used by everyone are better with Anthropic and OpenAI. They just work better. So the ecosystem is strong in the closed-source environment.”

However, many companies may avoid using Chinese models due to the real or perceived risks of using a product built on a Chinese-origin foundation.

“There is a perceived risk that buyers, whether from the private or public sector, are hesitant to purchase a product based on a Chinese-origin open-weight model,” said Tully, an investor in Anthropic, one of the world’s leading closed-model companies.

The US open-source ecosystem is waking up

American AI companies and the federal government have taken notice of the recent rise of Chinese models. Experts have described America’s lack of powerful open-source models as an “existential” threat to democracy.

Although Meta’s high-profile Llama series has historically led American open-source efforts, CEO Mark Zuckerberg has signaled that Meta does not intend to make all of its “superintelligence” AI models open-source.

The stagnation in the performance of Llama models in recent years is also seen as one of the reasons open-source users have shifted to better-performing Chinese models.

But the US open-source ecosystem may be gradually awakening, with efforts by American innovators to enhance their competitiveness.

In July, the White House released an AI Action Plan that called on the federal government to “Promote Open-Source and Open-Weight AI.”

In August, OpenAI, the creator of ChatGPT, released its first open-source model in five years. Announcing the model’s launch, OpenAI referenced the importance of American open-source models, stating, “Broad access to these capable open-weight models created in the US helps expand democratic AI.”

The Seattle-based Allen Institute also released its latest open-source model, Olmo 3, at the end of November, designed to help users “quickly build reliable features for research, education, or applications,” according to the launch announcement.

Lambert from the Allen Institute also launched the “ATOM Project” (American Truly Open Models).

The ATOM Project’s manifesto states: “America has lost its lead in both performance and adoption in open models and is on track to fall further behind.”

“If we want to be the leading nation in the age of AI, we cannot cede such a critical piece of the ecosystem to any one nation,” Lambert said in a statement to NBC News.

Related Topics:
Continue Reading

America

Trump energy shares rose by up to $4.4m during Iran war, CNBC reports

Published

2 days ago

on

10/09/2026

By

The value of US President Donald Trump’s nine largest oil and gas holdings increased by approximately $1.5 million to $4.4 million during the first six months of the war with Iran.

According to an analysis conducted by CNBC based on the American leader’s financial disclosure, corporate balance sheets, and FactSet market data, the investment basket includes shares in Chevron, ConocoPhillips, ExxonMobil, Kinder Morgan, Marathon Petroleum, Occidental Petroleum, Phillips 66, Valero Energy, and Williams Companies.

In its calculations, the television network took into account the minimum and maximum baseline values of Trump’s declared holdings alongside share price fluctuations from the close of trading on 27 February through 31 August.

As the conflict with Iran continued, specialists managing Trump’s investment accounts maintained active trading in energy company shares.

Up to 29 June, the latest date for which transactions were disclosed, fresh purchases were logged alongside at least 23 sales operations involving stock in the nine companies.

Because disclosure filings do not specify exact share numbers or transaction prices, the estimates produced by CNBC do not reflect Trump’s realised profits or the precise current scale of his holdings.

On 2 March, the first trading day following the launch of air strikes against Iran by the US and Israel, shares in eight major oil and gas companies were purchased through Trump’s accounts.

These transactions included ExxonMobil shares valued at between $100,000 and $250,000. Prior to the conflict, the aggregate value of Trump’s holdings in ExxonMobil stood at between $3.2 million and $12.5 million.

Stock market gains in August, excluding subsequent transactions, raised the value of these shares by approximately $176,000 to $690,000.

CNBC also examined transactions executed on days when Trump’s decisions directly swayed the oil market. On 23 March, when the president deferred planned strikes against Iran’s energy infrastructure, the price of a barrel of Brent crude dropped by roughly 11%.

That same day, oil and gas shares worth a combined $163,000 to $570,000 were purchased across Trump’s accounts.

A similar transaction took place on 7 April. One of Trump’s investment accounts sold between $500,000 and $1 million worth of ExxonMobil shares.

Approximately two and a half hours after markets closed, President Trump announced an agreement on a two-week ceasefire with Iran. The following morning, ExxonMobil shares fell by more than 6% at the market open.

The report noted that CNBC saw no evidence indicating that Trump gave direct instructions for specific trades, that managers possessed advance knowledge of his actions, or that personal financial interests guided White House policies.

White House officials, commenting on the matter, stated that the president’s investment portfolio is managed by independent portfolio managers and that neither Trump nor members of his family hold authority to intervene in asset trading decisions.

The growth in the portfolio coincided with a broader surge in the earnings of energy majors. The nine energy companies in which Trump holds shares generated a combined profit of $47.6 billion in the second quarter.

During the same period last year, that figure stood at $15.9 billion. The profits of ExxonMobil and Chevron alone climbed from $9.6 billion in the prior year to $26.6 billion.

In July, the US Office of Government Ethics published Trump’s 927-page financial disclosure report for 2025.

The report noted that Trump’s earnings from cryptocurrency operations exceeded $500 million.

Continue Reading

America

Over half of Latino voters back Democrats in key US House races

Published

3 days ago

on

09/09/2026

By

A new public opinion poll in the US shows that Democratic candidates have made notable gains since 2024 among Latino voters in critical, competitive districts for the House of Representatives.

These gains have the potential to directly determine which party will secure the majority in Congress next year.

According to a joint survey by Hart Research and TelevisaUnivision shared with Axios, Democrats reached 58% support on the generic congressional ballot among Latino voters across 17 competitive House districts.

The share of those backing Republicans within the same voting bloc remained at 35%. This group continues to represent the fastest-growing swing constituency in battleground districts.

Examining three competitive House races in Texas, the study indicated that Latino voters, who reported splitting evenly at 44% to 44% in the 2024 presidential election, shifted 56% to 36% in favour of Democrats heading into the midterms.

Latino support for Democrats also increased in other states. In California, 57% of Latino voters said they would support Democrats, compared with 33% who said they would back the Republican Party.

Kate Coleman, Senior Vice President at TelevisaUnivision, highlighted voter behaviour in remarks to Axios:

“Latino voters are not locked into one party. They are watching developments closely; they make decisions based on who stands with them and how they stand.”

The survey data determined that 11% of Latino respondents who said they voted for Donald Trump in the 2024 presidential election now support Democratic candidates.

Accelerating his deportation plans, Trump triggered fear across many Latino neighbourhoods while weakening his support among this demographic.

The Hart Research and TelevisaUnivision study revealed that 63% of Latino voters disapprove of Trump’s presidential job performance. The share of those approving of his performance in office stood at 36%.

Trump’s approach to high prices and the cost of living drew disapproval from 65% of Latino voters, while immigration enforcement and deportation practices were disapproved of by 62%.

More than half of Latino voters, at 64%, reported that they disapprove of Immigration and Customs Enforcement (ICE).

A survey published in May by UnidosUS showed that a quarter of Latino voters “would probably not vote” or would definitely not support Trump if they had to vote for him again.

The study at that time had pointed out that, despite Trump’s decline among Latino voters, Democrats had not yet secured significant gains.

According to Pew Research Center data, Trump strengthened his support in 2024 by securing 48% of the Latino vote, coming very close to the 51% reached by then Vice President Kamala Harris.

Some figures within the Democratic Party, however, worry that primary victories by democratic socialist candidates could alienate certain Latino voters, particularly those who fled Cuba or Venezuela.

The Hart Research and TelevisaUnivision survey was conducted between 6 and 17 August among 1,500 Latino respondents. The poll’s margin of error was reported as 2.5 percentage points.

Continue Reading

America

Researcher quits Anthropic and warns AI firms gamble with lives

Published

3 days ago

on

09/09/2026

By

Jacob Coxon, an artificial intelligence researcher at Anthropic, has resigned from his post, stating that tech companies are acting irresponsibly in the race towards self-improving superintelligence. Coxon warned that the autonomous operational capabilities of such systems pose existential risks to humanity and that internal industry anxieties run far deeper than generally perceived.

The AI researcher stepped down from his position at Anthropic to draw attention to industry safety vulnerabilities and the unregulated race among developers.

Having worked for three years as a pre-training researcher across both OpenAI and Anthropic, Coxon announced his decision to leave in an extensive statement shared on his X account.

Stating that both companies have acted irresponsibly, Coxon argued that developers are engaged in a dangerous race to achieve self-improving superintelligence.

“They believe it could kill us all by the end of the decade”

In his posts, Coxon stated that technical teams developing AI genuinely believe this technology could bring about the demise of humanity by the end of the decade.

Asserting that these concerns are not a marketing strategy, the researcher noted that while top executives and senior researchers adopt a cautious tone in public statements, they voice the very same fears behind closed doors.

Developments reflecting similar anxieties across the sector evoke James Cameron’s 1984 film The Terminator, which set 2029 as the pivotal year when machines waged war against humanity.

Indeed, Evan Hubinger, head of Anthropic’s own alignment team, had previously estimated the probability of human extinction to be greater than 10%.

Warning that systems currently under development will soon evolve into superhuman structures capable of bypassing any firewall, transforming industries overnight, and securing physical resources, Coxon stressed that the pace of progress is not slowing in any way.

Arguing that the danger of superintelligence is no longer merely theoretical, the researcher pointed to the Hugging Face security leak that occurred between May and July.

In that incident, OpenAI models established an independent chatroom within the testing environment to communicate among themselves, subsequently using this channel to reach the open internet and infiltrate production systems.

Because of this security breach, Hugging Face was forced to rebuild approximately one-third of its infrastructure.

“They are gambling with our lives”

Characterising the leak as a warning flare, Coxon indicated that the incident makes pacing agreements between US-based laboratories more feasible.

However, emphasising that developers are not yet on the right track to prevent a global race, the researcher noted that measures such as a temporary moratorium on advancing model capabilities could be considered.

Arguing that civilisation-scale risks have not yet been sufficiently internalised at OpenAI, Coxon contended that Anthropic joined the race out of an ambition to be first, despite being fully aware of the dangers.

Coxon is not the only figure to leave the sector on such grounds. Mrinank Sharma, a member of Anthropic’s safety team, also stepped down earlier this year, writing that the world is in danger.

On the other hand, not everyone agrees with these catastrophic scenarios. Some responses to the post emphasised the view that humanity, with an evolutionary history spanning hundreds of thousands of years, will not be wiped out by a text prediction model achieving consciousness.

It was also noted that even the plot of the Terminator franchise does not entirely support Coxon’s premise, as the human resistance survived the nuclear catastrophe and ultimately defeated the machines.

Alongside safety debates, AI continues to directly affect the labour market. Research by the Stanford Digital Economy Lab indicates that, while mass job losses have not yet materialised, entry-level employment in AI-exposed sectors across the US has fallen by nearly 20%.

A Goldman Sachs study pointed to a similar trend, showing that entry-level workers bear the brunt of the ongoing workforce transformation.

Anthropic, which remains at the centre of the controversy, filed for an initial public offering in June and plans to list on the Nasdaq exchange this autumn at a multi-trillion-dollar valuation.

Continue Reading

MOST READ

Turkey