Chinese developers scramble as OpenAI blocks access in China

At the recent World AI Conference in Shanghai, SenseTime showcased its new AI model, SenseNova 5.5, which boasts capabilities comparable to OpenAI's GPT-4o. With OpenAI blocking access to its tools in China starting July 9, Chinese AI companies like SenseTime are seizing the opportunity to attract new users. SenseTime is offering 50 million free tokens and free migration services to help developers transition from OpenAI to their platform. Other Chinese companies, including Baidu and Tencent Cloud, are also offering free tokens and migration services to capitalize on the shift.

The US decision to block OpenAI services in China comes amid rising tensions and export restrictions on advanced semiconductors crucial for AI development. This move has caused concern within China's AI community about equitable access to technology but has also opened doors for domestic companies to strengthen their foothold. Chinese AI companies are now in a race to capture the market, offering incentives to former OpenAI users and striving to close the performance gap with their US counterparts. Despite facing a price war that may impact innovation, this situation could accelerate the growth and testing of Chinese AI models.

Chinese commentators view OpenAI's departure as an opportunity for the country's AI industry to become more self-reliant. State media and academics argue that it could foster independence and innovation within China's technology sector. However, US restrictions are already affecting Chinese companies, with firms like Kuaishou experiencing limitations due to a chip shortage. As Chinese companies navigate these challenges, they may resort to creative solutions to circumvent sanctions and continue their AI advancements.

Hedonova has Outperformed the S&P 500 by Over 200%

Investing in alternative assets is complex. Hedonova simplifies this by offering access to a diverse portfolio, including startups, real estate, wine, art, and more. Since late 2019, it has outperformed the S&P 500 by 200%+. The SEC-regulated fund won the best multi-strategy hedge fund award in 2023.

This $1.5 Billion AI Company Is Building A ‘General Purpose Brain’ For Robots

Skild AI, a Pittsburgh-based robotics startup, has raised $300 million in a Series A funding round to continue developing its “general purpose brain” for robots. Founded in 2023 by former Carnegie Mellon professors Abhinav Gupta and Deepak Pathak, Skild AI aims to create a plug-and-play intelligence system that can be integrated into various robots. This innovative system allows robots to perform tasks such as climbing steep slopes, navigating obstacles, and identifying and picking up items. The funding round was led by Lightspeed Ventures, Softbank, Coatue, and Amazon founder Jeff Bezos, among others.

The cofounders leveraged their extensive research experience to build a massive database for training their AI model, which they claim is 1000 times larger than those of their competitors. The training methods included using human-operated robots, allowing robots to learn through trial and error, and incorporating millions of public videos. Pathak’s work on curiosity-driven learning has been pivotal, enabling robots to explore and learn from unpredictable outcomes. This comprehensive approach has led to robots demonstrating emergent capabilities, performing tasks they weren’t explicitly trained to do.

Despite facing competition from companies like OpenAI, Figure AI, and Covariant, Skild AI’s founders believe their vast data resources give them an edge. Their vision is to create a foundational model for robotics akin to what OpenAI has achieved with language models, potentially leading to artificial general intelligence for robots. Ken Goldberg, a robotics professor at UC Berkeley, acknowledges the challenges but sees potential in Skild AI’s data-centric approach. Investors like Stephanie Zhan from Sequoia Capital anticipate a significant breakthrough in the robotics field, similar to the advancements seen with digital intelligence.

Humane execs leave company to found AI fact-checking startup

Two top employees from Humane, Brooke Hartley Moy and Ken Kocienda, have left to start their own AI-driven fact-checking search engine called Infactory. Unlike Humane’s focus on AI hardware, Infactory will leverage AI to create a natural language interface, allowing users to easily obtain information from trusted sources without AI-generated hallucinations. The startup, which is in its early stages, will use a subscription model targeting enterprise clients like newsrooms and research facilities, emphasizing accuracy and reliability from data partners over content providers.

Infactory aims to provide a precise and efficient search experience, particularly useful for tasks such as comparing financial data between companies or sales figures for devices over several years. By avoiding AI in the actual search results and focusing on direct citations from reputable sources, Infactory plans to stand out in the market. The founders, now CEO and CTO, believe their extensive experience and selective partnership strategy will ensure the service's reliability and accuracy.

Despite their departure from Humane coinciding with the company's struggles post-launch, Hartley Moy and Kocienda assert their decision wasn't influenced by Humane's difficulties. They acknowledge the challenges of launching new hardware, respecting the courage and risk-taking involved. Infactory's launch is expected in the coming months, with the founders focusing on seed funding over the next year and a half to bring their vision to life.

Other stuff

• He created Oculus headsets as a teenager. Now he makes AI weapons for Ukraine 🔥🔥

• Beyond Language: Inside a Hundred-Trillion-Token Video Model 🔥

• It’s not just hype. AI could revolutionize diagnosis in medicine

• Alexa co-creator gives first glimpse of Unlikely AI’s tech strategy

• Turkish student arrested for using AI to cheat in university exam

• Andreessen Horowitz Is Building a Stash of More Than 20,000 GPUs to Win AI Deals 🔥

• Etsy adds AI-generated item guidelines in new seller policy

Hope you enjoyed today's newsletter

⚡️ Join over 200,000 people using the Superpower ChatGPT extension on Chrome and Firefox.

OR