Raphael Stevens

Melvin Chen moved to San Francisco from China to co-found AI design startup Lovart, which officially launched Wednesday.
The startup plans to focus on North America and doesn’t plan to operate in China for now, he said.
Analysts have long expected China’s AI advantage would likely lie in applications rather than models.

San Francisco-based AI design app Lovart officially launched Wednesday, with North American users in focus.

BEIJING — Chinese developers are powering some of the latest artificial intelligence tools aimed at a global market.
Melvin Chen moved to San Francisco from China to co-found AI design startup Lovart, which officially launched Wednesday — after claiming “800,000 users across 70 countries” for its test version.

“We will focus on North America as the first step,” Chen said in Mandarin, translated by CNBC. He previously led China operations for CapCut, a popular video-editing app from ByteDance that still ranks first in the photo and video category in Apple’s U.S. App Store.
Lovart uses AI to generate logos, stickers and other branding visuals based on text prompts. The new version launching Wednesday includes a “ChatCanvas” feature that claims to make specific edits easier — a client might ask a professional designer to switch two icons, a task difficult to explain only with words but simple when visuals are included, Chen said.
He expects Lovart to surpass 1 million users in the week after its launch. But he said the app isn’t coming to China soon, mostly because it’s based on Anthropic’s Claude 4 AI model and others from OpenAI — both of which aren’t officially available in China.
Beijing has to give generative AI models the green light for public use and operates a stringent firewall that blocks sites such as Google and Facebook. Companies also have their own rules about where their services can be used.

While most of Lovart’s team is based in San Francisco with the aim of better localizing the product, Chen said part of the production team is in China. He declined to share operating costs, and said the startup would seek investor funding after securing sufficient user growth.

Lovart has a free-to-use option, with monthly subscription fees of up to $90 for wider usage.

AI applications for video

In a global AI race, the U.S. government has in the last several years ramped up its restrictions on American companies selling advanced semiconductors to China. San Francisco-based OpenAI launched its ChatGPT chatbot in late 2022, and it wasn’t until January this year that China produced a clear rival with DeepSeek’s breakthrough.
But analysts have long expected China’s AI advantage would likely lie in applications rather than models, especially given that internet-based Chinese companies were able to build massive food delivery and short-video apps for the large local consumer market.
Already in AI video generation, Kuaishou’s Kling and Shengshu’s Vidu have gained global users in the last 18 months. In the realm of AI agents that can automatically perform a series of complex tasks, Manus has caught international attention.
“China-affiliated teams are increasingly influential, driven by concentrated technical talent, agile development culture, and policy support for AI commercialization,” said Charlie Dai, vice president and principal analyst at Forrester. “They excel in cost-efficient model training and rapid consumer app iteration, often prioritizing open-source accessibility.”
“Chinese models now compete globally, challenging U.S. dominance while lowering AI costs,” he said.
Another advantage is that China models such as DeepSeek and several others are open source, meaning they are free for developers to download and use.
Hugging Face, an online platform that allows people to try out open source AI models, regularly show that China models are among the top trending ones for users.
As of Wednesday, the Kimi K2 coding-focused model that was launched this month ranked first on the site, followed by Alibaba’s Qwen3 coding-focused modes that launched earlier in the day. In image-to-3D models, Tencent’s Hunyuan ranks first, while France-based Mistral’s Voxtral ranks first in audio-text-to-text.
Chen said Lovart will focus on AI for generating images and videos rather than 3D models.
“AI is the new camera … [for] capturing human imagination,” he said. He said the startup aims to build traction by holding events with the design community, including in New York, Tokyo and Europe.

Weekly analysis and insights from Asia’s largest economy in your inboxSubscribe now

ChatGPT is by far the most popular AI app in the West, with 70 million monthly users on average in the U.S. and 144.6 million in Europe as of July, according to Sensor Tower.
Google’s Gemini was a distant second in both markets, but while Microsoft Copilot ranked third in the U.S., DeepSeek held the third spot in Europe, the data showed.
During a visit to Beijing last week, Nvidia CEO Jensen Huang said nearly all of DeepSeek’s users had downloaded the model to run it locally in countries around the world. He also emphasized that priorities for AI development are shifting.
“I think over time it will be increasingly less important which one of the models are the smartest,” he said. “It’s going to be which one of the models are the most useful.” More

Capital One shares rose on Tuesday evening despite the company reporting an extremely noisy second-quarter result due to the Discover integration. Still, we like where the company is headed with this game-changing acquisition. Revenue in the three months ended June 30 increased 31% year over year to $12.5 billion, missing the consensus estimate of $12.7 billion, according to LSEG. Adjusted earning per share (EPS) increased 75% year over year to $5.48, exceeding the $3.72 estimate, LSEG data showed. Shares are trading up about 3% in extended trading Tuesday night to around $224 per share. If the stock closes above $220.91 on Wednesday, it will mark a new all-time high. Bottom line This was not the easiest quarter to judge, but long-term benefits of owning Discover are easy to see. The blockbuster Discover acquisition, which closed on May 18, required a lot of different accounting treatments and analyst estimates were all over the board. For example, Capital One actually reported a quarterly net loss of $4.3 billion, or $8.58 per share, based on Generally Acceptable Accounting Principles (GAAP) — but, on an adjusted basis to strip out one-time impact from the deal, the company turned a huge profit of $5.48 per share. One of the largest financial impacts from the deal was the $8.8 billion worth of initial allowance build for Discover’s non-purchased credit deteriorated loans. The accounting treatment for Discover’s book of business is why there was a significant increase in the reported companywide provision for credit losses. Provisions for credit losses are funds that Capital One sets aside to cover potential loan defaults; the higher the provisions, the worse sign of credit quality. Backing out the Discover provisions tells a different story. If it was still a standalone company, Capital One would have had an allowance release of around $900 million, which is a great sign of improving credit trends. This is a big difference, to say the least. Capital One Financial Why we own it : Capital One’s acquisition of Discover is a transformative deal with significant strategic advantages and financial benefits. There are also several billions of dollars worth of expense and network synergies that should make this deal highly accretive to earnings per share. Lastly, the acquisition strengthens Capital One’s balance sheet, allowing for aggressive share repurchases in the future. Competitors : American Express, MasterCard, Visa Most recent buy : May 23, 2025 Initiated : March 6, 2025 Beyond the nitty gritty of the credit metrics, the focus of Tuesday night’s earnings call was all about the Discover integration and what management’s plans are now that it owns a payments network — the most coveted part of the $35 billion acquisition. As CEO Richard Fairbank proudly pointed out, “There are only two banks in the world with their own network, and we are one of them. We are moving to capitalize on this rare and valuable opportunity.” American Express is the other. Our thesis is that the Discover acquisition will boost Capital One’s earnings power and expand its price-to-earnings multiple. With the integration just getting started, the stock remains undervalued. Although Capital One will have to invest aggressively to achieve its vision, those returns should be worth the costs and help the company grow sustainably for years. We’re reiterating our buy-equivalent 1 rating and price target of $250. Deal outlook On the earnings call, the company provided some early thoughts on the how Discover integration is progressing. Broadly speaking, the integration “is off to a great start,” and that’s good to hear since so much of our thesis hinges on this deal being a success. However, management now expects integration costs to be “somewhat higher” than its previous announced target of $2.8 billion, which is a slightly negative development. According to Fairbank, the “integration budget” covers expenses like deal costs; moving Discover onto Capital One’s tech stack; integrating products and experience; additional investments in risk management and compliance; integrating talent; and taking care of employees. In addition to the higher cost outlook, the phrase “sustained investment” came up multiple times on the conference call. Fears of endless spending to make the deal work could spook some investors. However, the firm believes these sustained investments will lead to sustained growth and stronger returns for the long run. “The portfolio of opportunities we have is the broadest and biggest set of opportunities that I’ve seen in our history. But the only way to get there is with investment,” Fairbank said — and we’re banking on Fairbank being right. “I think there’s a lot of value creation opportunity, but we’re going to invest significantly to get there,” he later added. On the synergy side, Capital One said it’s on track to hit its target of $2.5 billion of net synergies, which is made up of cost savings and revenue synergies generated by moving its debit business and some of its credit business onto the Discover network. Capital One began the process of reissuing Capital One debt cards onto that network last month, Fairbank said. The conversion process will continue “in phases through early 2026,” he said. Longer term, the company sees a significant opportunity to invest in the network to achieve greater international acceptance and build a global network brand. Management wants to do this to lure bigger spenders onto the Discover network, and doing so could eventually could help the company exceed its synergy targets , Fairbank has said. Commentary As mentioned earlier, the actual quarterly results were hard to evaluate versus expectations because the estimates themselves varied tremendously. Analysts need time to fine-tune their models for the combined company. For that reason, we’re not putting too much stock into all the red seen in the chart above. The bearish view on Capital One is that the tariff-driven plunge in consumer sentiment would hurt the economy and materially impact Capital One’s credit performance. Since Capital One is one of the more exposed credit card companies to subprime, it’s usually the first to feel the pain of an economic slowdown. And yet, the bank’s credit performance has been healthy and steadily getting better. “Capital One’s card delinquencies have been improving on a seasonally adjusted basis since October of last year, and our losses have been improving since January of 2025,” Fairbank said on the call. Capital One’s “legacy” domestic card portfolio, which does not include Discover, also saw its net charge off rate decline 55 basis points year over year to 5.5%. Net charge-offs refer to the amount of debt a bank has written off as uncollectible, minus any recoveries. A decline is a good thing. Toward the end of Tuesday night’s call, Fairbank spoke more generally about the health of the U.S. consumer and economy, striking an upbeat tone. “If we don’t read the news and just look at what our customers are telling us with their behaviors, it is a picture of strength,” he said. As for buybacks, the company repurchased $150 million worth of stock in the quarter, bringing its full-year total to $300 million. Following another successful round of Federal Reserve stress tests in June, there’s a lot of potential here for years of multibillion dollar buybacks. But management is still working through the internal modeling of the combined company, and they plan on making an update once that is complete. (Jim Cramer’s Charitable Trust is long COF. See here for a full list of the stocks.) As a subscriber to the CNBC Investing Club with Jim Cramer, you will receive a trade alert before Jim makes a trade. Jim waits 45 minutes after sending a trade alert before buying or selling a stock in his charitable trust’s portfolio. If Jim has talked about a stock on CNBC TV, he waits 72 hours after issuing the trade alert before executing the trade. THE ABOVE INVESTING CLUB INFORMATION IS SUBJECT TO OUR TERMS AND CONDITIONS AND PRIVACY POLICY , TOGETHER WITH OUR DISCLAIMER . NO FIDUCIARY OBLIGATION OR DUTY EXISTS, OR IS CREATED, BY VIRTUE OF YOUR RECEIPT OF ANY INFORMATION PROVIDED IN CONNECTION WITH THE INVESTING CLUB. NO SPECIFIC OUTCOME OR PROFIT IS GUARANTEED. More

President Donald Trump continued his barrage of criticism against the Federal Reserve but seemed to take a step back from any lingering plans to fire Chair Jerome Powell. “He’s going to be out pretty soon anyway,” Trump told reporters.
Trump has accused Powell of being political and nicknamed him “too late” when it came time to adjusting interest rates.

President Donald Trump on Tuesday continued his barrage of criticism against the Federal Reserve but seemed to take a step back from any lingering plans to fire Chair Jerome Powell.
“I think he’s done a bad job, but he’s going to be out pretty soon anyway,” Trump told reporters during a White House exchange. “Eight months, he’ll be out.”

The exchange came amid continued speculation over Powell’s job security and legal questions over what authority Trump has to remove the central bank leader.
Going back to his first term in office when he appointed Powell, Trump has repeatedly criticized the Fed for an at-times cautious approach to cutting, though it lowered its benchmark borrowing rate a full percentage point in late 2024 around the time of the presidential election. Powell’s term as chair ends in May 2026.
Trump has accused Powell of being political and nicknamed him “too late” when it has come time to adjusting rates.
“He’s too late all the time. He should have lowered interest rates many times,” Trump said. “People aren’t able to buy a house because this guy is a numbskull. He keeps the rates too high, and [is] probably doing it for political reasons.”
The president has floated the idea of trying to fire Powell, going so far as to ask Republicans gathered in his office last week about the idea. However, he soon after called it “highly unlikely” that he would try.

In addition to the criticism over rates, the White House has called into question a $2.5 billion renovation project the Fed has undertaken on two of its Washington, D.C., buildings. Administration officials plan to tour the site Thursday, according to a post Tuesday on X from James Blair, the White House deputy chief of staff.
Treasury Secretary Scott Bessent, who earlier in the day Tuesday said Powell “has been a good public servant,” said at the White House that the Fed needs to “do a big internal investigation to understand not their monetary policy but everything else. The Fed has had this big mission creep, and that’s where a lot of the spending is going.”
Trump complained that the Fed “can be spending $2.7 billion to build a building. They don’t do anything.”

Don’t miss these insights from CNBC PRO More

For years, America’s job market has rewarded the footloose. The surest route to a higher salary, the usual advice goes, is to string together a series of one- or two-year stints, each paying a bit better than the last. Career gurus on TikTok set videos of their own salary progression to jaunty pop beats, cloaking online bragging as guidance for the uninitiated. On Reddit, posters debate just how little time in a role a job-hopper can get away with before future employers might start to fret about disloyalty. (A year or so is the consensus, though a brave few argue for six months.) More

For years, America’s job market has rewarded the footloose. The surest route to a higher salary, the advice goes, is to string together a series of one- or two-year stints, each paying a bit better than the last. Career gurus on TikTok set videos of their own salary progression to jaunty pop beats, cloaking online bragging as guidance for the uninitiated. On Reddit, posters debate just how little time in a role a job-hopper can get away with before future employers might start to fret about disloyalty. (A year or so is the consensus, though a brave few argue for six months.) More