Meta Scientist LeCun Launches AI Startup as Google Invests $90 Billion

Yann LeCun, Meta's chief AI scientist and a pioneer in the field, is departing the company at the end of the year to launch his own AI startup. His new venture will focus on "world models" AI, aiming to create systems that understand the physical world, remember, reason, and plan complex actions. LeCun, who co-founded Meta's FAIR AI research division in 2013, has often expressed skepticism about Meta's primary focus on large language models, preferring his JEPA approach. Despite his departure and differing views, Meta plans to partner with LeCun's new company, aligning some of its research with Meta's business goals. This move comes as Meta's AI organization recently reorganized under Superintelligence Labs. The broader AI market is currently under intense scrutiny, with Nvidia's upcoming earnings report poised to test whether the current AI boom is sustainable or an "AI bubble." The "Magnificent 7" tech companies, including Nvidia, Google, Apple, and Meta, have driven most of the S&P 500's gains, with a combined market value of $15 trillion. This market concentration is even higher than during the 1999 dot-com bubble. Companies like OpenAI show much lower revenues compared to their massive spending and high valuations, raising concerns that global stocks could face a severe downturn if AI revenues do not catch up with investments. Google CEO Sundar Pichai recently showcased the company's Tensor Processing Unit (TPU) lab in California, highlighting Google's commitment to developing specialized chips for its AI operations. Google is investing over $90 billion yearly in AI, a three-fold increase in four years, despite warnings about stretched valuations from entities like the Bank of England and OpenAI CEO Sam Altman. Meanwhile, Nvidia is collaborating with Menlo Micro to significantly speed up the testing process for its AI chips, improving GPU testing by 30% to 90% faster using new switching chips. Significant investments are flowing into AI infrastructure globally. Brookfield is partnering with Nvidia and the Kuwait Investment Authority to launch a $100 billion global AI infrastructure program. Separately, GMI Cloud and Nvidia are building a $500 million AI Factory in Taiwan, intended as a major hub for AI development across Asia. The U.S. economy is also increasingly relying on AI growth, with President Donald Trump highlighting its role in stock market gains. Saudi Crown Prince Mohammed bin Salman pledged $1 trillion in investments with U.S. companies, largely aimed at transforming Saudi Arabia into a major AI data hub. In other developments, Target Corp. announced a partnership with OpenAI to integrate AI technology, hoping to reverse a sales slump and enhance personalized shopping, supply chains, customer service, and product discovery. However, Microsoft issued a warning regarding an experimental AI feature in Windows, Copilot Actions, stating it could infect devices and steal sensitive user data due to known flaws like "hallucinations" and "prompt injection" attacks. Microsoft advises only experienced users enable this beta feature. Addressing security concerns, U.S. Senators Todd Young and Mark Kelly introduced the Advanced AI Security Readiness Act, which would require the NSA to create a playbook to guard against foreign threats to American AI systems and improve cyber attack responses.

Key Takeaways

• Yann LeCun, Meta's chief AI scientist, is leaving to launch a new AI startup focused on "world models" AI, aiming for systems that understand the physical world, remember, reason, and plan complex actions.
• Meta will partner with Yann LeCun's new venture, despite his past disagreements with Meta's focus on large language models.
• Nvidia's upcoming earnings report is a critical test for the AI market, with concerns about an "AI bubble" due to high valuations and lower revenues from companies like OpenAI compared to their investments.
• Google is investing over $90 billion yearly in AI, a three-fold increase in four years, and is developing its own Tensor Processing Unit (TPU) chips in California.
• The "Magnificent 7" tech companies, including Nvidia, Google, Apple, and Meta, hold a combined market value of $15 trillion and make up one-third of the S&P 500, showing higher market concentration than during the 1999 dot-com bubble.
• Brookfield, Nvidia, and the Kuwait Investment Authority are launching a $100 billion global AI infrastructure program, while GMI Cloud and Nvidia are building a $500 million AI Factory in Taiwan.
• Saudi Crown Prince Mohammed bin Salman pledged $1 trillion in investments with U.S. companies, largely to establish Saudi Arabia as a major AI data hub.
• Target Corp. partnered with OpenAI to integrate AI technology, aiming to boost sales through personalized shopping, improved supply chains, and enhanced customer service.
• Microsoft warned that its experimental AI feature, Copilot Actions in Windows, could infect devices and steal sensitive user data due to vulnerabilities like "prompt injection" attacks.
• U.S. Senators introduced the Advanced AI Security Readiness Act, requiring the NSA to create a playbook to protect American AI technology from foreign threats.

Yann LeCun leaves Meta to launch AI startup

Yann LeCun, Meta's chief AI scientist and a pioneer in AI, is leaving the company to start his own venture. His new startup will focus on "world models" AI, aiming to create systems that understand the physical world, remember, reason, and plan complex actions. Meta will partner with LeCun's new company, despite recent changes in Meta's AI unit and a different approach to AI development. LeCun, 65, co-founded Meta's FAIR AI research division in 2013 and won the Turing Award in 2019.

AI pioneer Yann LeCun exits Meta for new company

Yann LeCun, Meta's chief AI scientist, announced he will leave his role at the end of the year to form a new startup. This company will research advanced AI that can understand the physical world, remember things, reason, and plan complex actions. Meta will partner with his new venture. LeCun joined Facebook in 2013, co-founded its AI research division, and also works as a part-time professor at New York University.

Yann LeCun departs Meta after 12 years for AI startup

Yann LeCun, Meta's chief AI scientist, is leaving after 12 years to launch his own AI startup. His departure comes during a period of change within Meta's AI organization, which recently reorganized under Superintelligence Labs. LeCun has often disagreed with Meta's focus on large language models, preferring his JEPA approach to help AI understand the physical world. His new company will pursue this different vision for AI development.

Meta AI chief Yann LeCun starts new company

Yann LeCun, Meta's chief AI scientist, announced he will leave his job at the end of the year. He plans to start a new company to research advanced AI that can understand the physical world, remember, reason, and plan complex actions. Meta will partner with LeCun's new startup, and some of its research will align with Meta's business goals.

Yann LeCun leaves Meta to launch AI research firm

Yann LeCun, Meta's chief AI scientist, will leave at the end of the year to start a new AI research company. His startup will focus on advanced AI that can understand the physical world, remember, reason, and plan complex actions. Meta will partner with his new company, even though LeCun has been skeptical of large language models and advocates for open-source AI. LeCun co-founded Meta's AI research division in 2013 and won a top computer science prize in 2019.

Nvidia earnings to test AI market bubble fears

Nvidia's upcoming earnings report is very important for the global stock market and will help answer if the AI boom is a bubble. The "Magnificent 7" tech companies, including Nvidia, have driven most of the S&P 500's gains, investing huge amounts in AI. However, AI companies like OpenAI show much lower revenues compared to their spending and high valuations. If AI revenues do not catch up with these massive investments, global stocks could face a severe downturn. The current market concentration in a few tech firms is even higher than during the dot-com bubble of 1999.

Google's AI chip factory fuels tech race

Google CEO Sundar Pichai showcased the company's Tensor Processing Unit lab in California, where special chips called TPUs are being developed to power all Google's AI. Google is investing over $90 billion yearly in AI, a three-fold increase in four years, despite concerns about an "AI bubble." The Bank of England and OpenAI CEO Sam Altman have also warned about stretched valuations. Five major tech companies, including Google, Nvidia, Apple, Meta, and OpenAI, hold a combined market value of $15 trillion. The "Magnificent 7" now make up one-third of the S&P 500, showing a higher market concentration than during the 1999 dot-com bubble.

Trump links US economy to AI growth and Saudi investment

President Donald Trump is increasingly relying on the tech sector and AI to boost the U.S. economy. This was clear during Saudi Crown Prince Mohammed bin Salman's visit, where he pledged $1 trillion in investments with U.S. companies. These funds will largely help Saudi Arabia become a major AI data hub. Trump highlighted AI's role in recent stock market gains and new investments, which have helped offset economic uncertainty. However, experts warn that AI companies' reliance on debt could lead to a vulnerable phase next year. Elon Musk predicted AI would eliminate poverty, while Nvidia's Jensen Huang noted that jobs will change.

Trump's economy relies on AI and Saudi investment

President Donald Trump is increasingly depending on the tech sector and AI to drive the U.S. economy. This was highlighted during Saudi Crown Prince Mohammed bin Salman's visit to the U.S. The Crown Prince pledged to invest $1 trillion with American companies. This large investment aims to transform Saudi Arabia into a major AI data hub, using its oil and natural gas resources.

Brookfield Nvidia Kuwait fund launch 100 billion AI program

Brookfield is partnering with Nvidia and the Kuwait Investment Authority, a sovereign wealth fund, to launch a massive $100 billion global AI infrastructure program. This collaboration aims to build essential facilities needed for artificial intelligence development worldwide. The significant investment highlights the growing demand for robust AI infrastructure.

GMI Cloud and Nvidia build 500 million dollar AI factory

GMI Cloud announced a new partnership with NVIDIA to build a next-generation AI Factory in Taiwan. This project involves a US$500 million investment. The facility will be a key part of NVIDIA's global AI Factory strategy. It will serve as a major hub for AI development across Asia, boosting the region's AI infrastructure.

Target partners with OpenAI to boost sales

Target Corp. announced a partnership with OpenAI, the creator of ChatGPT, to help reverse its sales slump and gain an edge in the AI market. This strategic move follows a quarter of disappointing results for the retailer. While details are still secret, Target CEO Brian Cornell suggested AI could transform retail through personalized shopping and better supply chains. The company hopes integrating OpenAI's technology will improve customer service, product discovery, and inventory management, driving future growth.

AI market faces reality check with Nvidia earnings

The AI market is facing a critical "moment of truth" as investors question if the current boom is sustainable. Nvidia's upcoming earnings report is key, as it will show if the demand for AI chips is real or just a cycle of companies buying GPUs to appear innovative. The S&P 500 has already slipped, and investors are now focusing on actual earnings and valuations. While AI technology is transformative, many companies struggle to show how their large AI investments will generate profits. The risk is that investors may lose patience, similar to the dot-com bubble, if real financial benefits do not appear soon.

Nvidia and Menlo Micro speed up AI chip testing

Nvidia is working with Menlo Micro to significantly speed up the testing process for its AI chips. Nvidia sells millions of AI chips, which need thorough testing on special circuit boards. Older testing chips struggled with the high power and speed of modern AI chips. Menlo Micro, a startup, developed new switching chips using tiny metal switches. This technology improves test board performance, allowing Nvidia to test its GPUs 30% to 90% faster. Other major chipmakers are also starting to use Menlo Micro's innovative testing chips.

Microsoft warns AI feature can steal data

Microsoft warned that an experimental AI feature in Windows, called Copilot Actions, could infect devices and steal sensitive user data. Critics are concerned that Big Tech releases new features before fully understanding their risks. The warning stems from known flaws in large language models, such as "hallucinations" and "prompt injection" attacks. These vulnerabilities could allow malicious content to override AI instructions, leading to data theft or malware installation. Microsoft suggests only experienced users enable Copilot Actions, which is currently in beta. However, security experts doubt the effectiveness of these warnings, comparing them to past ineffective warnings about dangerous software macros.

Senators introduce bill for AI security playbook

U.S. Senators Todd Young and Mark Kelly introduced the Advanced AI Security Readiness Act to protect American AI technology. This new law would require the NSA to create a playbook to guard against foreign threats to AI systems. The act aims to find weaknesses in AI, improve how the country responds to cyber attacks, and keep America's lead in technology. Similar legislation is also being considered in the House of Representatives.

Sources

• Meta chief AI scientist Yann LeCun is leaving to create his own startup
• Meta's chief AI scientist Yann LeCun to leave Meta and start new AI research company
• Yann LeCun, Meta's chief AI scientist, is walking away after 12 years to start his own AI startup
• Meta's chief AI scientist Yann LeCun to leave Meta and start new AI research company
• Meta’s chief AI scientist Yann LeCun to leave Meta and start new AI research company
• Nvidia's earnings could answer the AI bubble question and upend global markets in moment of truth for Magnificent 7
• The Silicon Valley 'AI factory' at the heart of the tech race
• Trump’s economy depends on AI for growth, a reality made clear in Saudi crown prince’s US visit
• Trump's economy depends on AI for growth, a reality made clear in Saudi crown prince's US visit
• Brookfield partners with Nvidia, Kuwait wealth fund for $100 billion AI infrastructure program
• GMI Cloud Launches Next-Generation AI Factory in Taiwan
• Target announces partnership with OpenAI as it aims to reverse sales slump
• The AI Trade's Moment of Truth
• Nvidia, Menlo Micro collaboration speeds up AI chip testing
• Critics scoff after Microsoft warns AI feature can infect machines and pilfer data
• Press Release: Senators Todd Young and Mark Kelly Introduce Advanced AI Security Readiness Act