Despite SpaceX reporting higher-than-expected revenue from artificial intelligence expenditures in its first public report, investors remain concerned about how long the profitable Starlink business can cover the costs of data centers and Nvidia chips.
The company's shares fell by approximately 12% on Wednesday, dropping significantly below the $135 IPO price in less than two months after the high-profile debut. Market analyst Carolane de Palmas from the brokerage firm ActivTrades noted that with high capital expenditures persisting and investor enthusiasm cooling, the stock could remain under pressure until the lock-up period expires.
She added that this could lead to significant volatility as markets reassess SpaceX's valuation and cash burn trajectory. SpaceX reported AI revenue that tripled compared to the previous year and announced several new cloud computing agreements, even though quarterly AI capital expenditures reached $15.8 billion.
CFO Brett Johnson stated that the economics of these investments are rapidly improving but also noted that AI spending will remain high. Johnson specified that the current economics allow for the payback of new compute capacity capital investments in less than a year, and also mentioned that SpaceX signed another $6.7 billion in cloud computing contracts since the end of the second quarter and is on track to achieve $100 billion in annual revenue by the end of the year.
This contrasts with traditional data center investments, which typically require several years to recoup initial costs. David Wagner, portfolio manager at Aptus Capital Advisors, observed that 'Elon continues to surprise investors with what innovations and technology can achieve, but there has always been a mismatch in the timeframe for when this execution will happen,' referencing Musk's often optimistic forecasts for his EV company Tesla, which he frequently failed to meet.
He continued: 'I believe the numbers. I would say yes, these numbers are aggressive, but it is not fantasy. The details exist; they just require flawless execution.'
Comments during the earnings call marked a shift in investment debates surrounding SpaceX. Before the report's publication, investors primarily viewed Starlink's growing cash flows as the main source of funding for the company's AI ambitions. Now, management asserts that the AI infrastructure itself is beginning to generate sufficient revenue for further expansion.
Nevertheless, the company spent about $18.4 billion on capital expenditures during the quarter, accounting for roughly one-fifth of the $85.7 billion raised in the June IPO, and remained deeply negative on free cash flow while continuing to heavily invest in AI infrastructure.
Josh Gilbert, lead analyst at trading platform eToro, commented: 'We have seen the same intense scrutiny on large tech companies this earnings season, where investors questioned open wallets and began demanding visible returns. SpaceX faces this test with added complexity because it asks shareholders to fund data centers in orbit.'
SpaceX's AI business generated $2.6 billion in revenue in the second quarter, more than three times that of the previous year, although it remained operationally unprofitable. Nevertheless, the company does not plan to reduce its spending. Johnson stated that capital expenditures over the next two quarters are likely to remain at a level similar to the second quarter, as SpaceX continues to expand AI compute power, Starship production, and next-generation Starlink satellites.
Michael Monaghan, portfolio manager of the Founders 100 ETF holding SpaceX stock, noted: 'The new compute power is monetizing so quickly that it behaves more like a cost of goods sold than a capital expenditure.'
SpaceX executives stated that demand for AI computing continues to outstrip supply, and they expect to complete the year with over 2 GW of computing power. If demand remains strong enough to fully utilize this infrastructure, AI will increasingly be able to fund its own growth instead of relying on cash generation from Starlink.
Drew Capes, portfolio manager at Polen Capital, which holds a position in SpaceX, concluded: 'The link between capital expenditures and revenue is unstable, so capital expenditures must decrease, or revenue must grow tremendously, and this is where belief in Musk's vision, engineering leadership, and track record of execution differentiates bulls from bears.'
The team of Gemini leaders at Google has left the company due to a major overhaul of leadership within the artificial intelligence division. Among those departing are engineer Jeff Dean and other leaders of the Gemini model.
These changes occur at a critical moment for Google DeepMind: the flagship version of the latest Gemini model has yet to be released, despite a planned June launch. This raises concerns among investors and industry circles that Google is falling behind competitors Anthropic and OpenAI, who recently recruited leading specialists from Google AI.
Following this news, shares of Google's parent company, Alphabet, dropped by 4%. Hassabis, a Nobel laureate, announced in a memo to employees his transition from CEO of Google DeepMind to Chairman of the Board, accepting the new role of Chief Scientist at Alphabet. He explained the change in role by citing the proximity to achieving Artificial General Intelligence (AGI), which is an advanced state of AI where computers essentially become as intelligent as humans.
Sundar Pichai, head of Alphabet, noted in a separate corporate memo that he had long discussed with Hassabis a role that would allow him to fully concentrate on shaping the future of AGI, calling this work vital for Alphabet and humanity.
Korey Kavukcuoglu, the division's CTO, will take over daily operational duties. He will assume the position of Senior Vice President of Business Units, in addition to his parallel role as Alphabet's Chief AI Architect.
Dean, Sanjeev Ghemawat, Oriol Vinyals, and Quoc Le—some of the most renowned and highly regarded leaders in AI and engineering at Google—have left the company to found a startup called Discovery Loop. This startup will specialize in research breakthroughs in machine learning, science, and engineering, structured as a public benefit corporation.
Alphabet has not provided detailed information regarding the reasons for these departures or any overlap in timelines. As Reuters reported in a report last year, some employees who worked with Hassabis claimed that he often prioritized scientific research over generating profit for Google. According to three sources familiar with the process, the London-born scientist viewed the Nobel Prize as a business objective for DeepMind. Previously, insiders reported that he resisted other initiatives that could bring new revenue to Alphabet or improve the company's standing in the AI race.
Allies within Google dismissed the idea that he was too focused on long-term science or neglecting business.
Leaders of Google Cloud, the company's primary source of AI revenue, welcomed Kavukcuoglu's appointment on Wednesday, viewing it as positive news for advancing commercialization efforts. A source with direct knowledge of the situation stated that Kavukcuoglu had already been involved in collaborations between DeepMind and Cloud more actively than Hassabis, and his appointment as head of the AI lab was intended to steer Gemini toward a more product-oriented direction.
An Alphabet representative stated that Hassabis would focus on research and strategy related to the social impact of AGI and would have few direct reports.
In his memo, Hassabis noted 'great progress' in Google's AI models, including the raw update named Gemini 4. He also mentioned that he would dedicate more time to Isomorphic Labs, a drug discovery company spun out of DeepMind, which he founded and continues to lead.
He emphasized: 'I have always believed that the best application of AI should be improving human health. The time has come for AI to prove its undeniable value to the world, and what better way to demonstrate that than by finally helping cure diseases like cancer.'
Dean and Hassabis were long considered two leading figures in AI at Google before the field became a hub of technological innovation. Hassabis co-founded DeepMind in 2010 and continued to lead the London lab after its sale to Google for approximately $650 million in 2014.
Dean co-founded Google Brain in 2011, a US-based division that researched AI separately from Hassabis's team. By then, he was already a respected figure within Google, known for his collaboration with Ghemawat on implementing many landmark infrastructure projects that still support the company's operations.
In 2023, after OpenAI's launch of ChatGPT caught Google off guard, Pichai announced the merger of DeepMind and Google Brain into a new organization called Google DeepMind. Hassabis was appointed its CEO, and Dean was named Chief Scientist.
Until recently, Dean, Vinyals, and Noam Shazeer jointly led the development of Gemini. Under their leadership, Google released Gemini 3 last November, which the industry widely perceived as narrowing the gap with leading models from Anthropic and OpenAI.
However, subsequent achievements by competitors, along with the rise in popularity of cheaper open-source Chinese models, placed Google in a defensive position. At Google's annual developer conference in May, Pichai and other Google leaders emphasized Gemini's price advantage rather than its cutting-edge capabilities.
Shazeer left for OpenAI in June, less than two years after Google spent billions to hire him and his research team from the startup he led. In the same week, senior researcher John Jumper, who won the Nobel Prize with Hassabis in 2024, announced his move to Anthropic.
Vinyals, Le, and former OpenAI chief scientist Ilya Sutskever co-authored a 2014 research paper that became foundational for modern AI research, establishing the concept that scaling data for AI system 'pre-training' would elevate them to new heights.
Brian Mallory, Chief Market Strategist at Zacks Investment Management, commented: 'Google possesses a deep talent pool, but the loss of these four individuals will leave a mark for some time. Restoring innovation and execution will not be easy, and losing key players will not make the situation any better.'
Ola Electric has announced a structural change in its market entry strategy, opening its sales and service networks to dealer partners across India. This decision comes five years after the company launched its first electric scooters.
Previously, the company's growth was driven by its own stores, which were used to raise awareness about electric vehicles, strengthen the brand, and build the largest base of two-wheeler EV customers in India, exceeding one million users. However, as the EV industry matures and consumer adoption increases, the company is revising its operating model.
In the coming months, the brand stores will be reoriented to provide brand and product experience, while dealer partners will become the foundation for local sales, servicing, and scaling nationwide. This move followed a month of active engagement with dealers across the country.
Ola Electric noted that the market reaction has been strong. Dealers expressed confidence that the company's product portfolio, combined with their ability to execute plans locally, can significantly accelerate both Ola's sales and the overall adoption of EVs in India.
Bhavish Aggarwal, Chairman and Managing Director of Ola Electric, stated: 'This is a major shift in our market approach, reflecting the direction of the EV industry in India. We initially built our own stores because we needed to win trust in EVs store by store, city by city. This phase laid the foundation.'
He added that now that the category is mature, the company's task is to deliver EVs to every corner of the country, and dealer partners possess an understanding of their local markets that cannot be replicated by a proprietary network at this scale. The brand stores will now serve as places where customers can receive the best brand and product experience, while dealer partners will disseminate this experience across all cities and regions of India.
For Ola Electric's dealer partners, the company offers a strong brand established over five years in the market, significant opportunities for regular servicing based on the largest base of two-wheeler EV owners in India—over 10 lakh customers, a full product range including scooters, motorcycles, and energy products like 'Ola Shakti', as well as planned releases of additional energy products, and an economic model for dealers designed to ensure high returns for partners.
Ola Electric expects this program to drive tangible local growth by Diwali 2026. As part of this transition, BVR Subbu will rejoin Ola Electric as a Senior Advisor to support the project's implementation. Subbu, former President of Hyundai Motors India, previously served on the Ola Electric board of directors and mentored Aggarwal.
The Common Facility Center (CFC) scheme of the Uttar Pradesh government is a major part of the One District One Product (ODOP) program. The objective of this scheme is to develop advanced common technical infrastructure for artisans, micro, small, and medium enterprises (MSMEs), producer groups, and other stakeholders associated with district-specific ODOP products.
This scheme aims to strengthen the value chain of traditional industries and improve the availability of modern infrastructure for enterprises working in ODOP product clusters by providing access to shared facilities.
A Common Facility Center is a shared infrastructure facility prepared for enterprises related to the manufacturing of identified ODOP products. Instead of building separate infrastructure for each enterprise, this scheme provides an opportunity for multiple users of the same product cluster to utilize shared facilities, aiding in production, quality improvement, product development, and marketing.
These centers are established on a group-based method and are implemented through a Special Purpose Vehicle (SPV). This SPV can be a registered society, trust, producer company, cooperative society, or any other legal entity recognized under the scheme.
The facilities developed under CFC can vary according to the approved projects and the requirements of the identified ODOP product. According to the scheme, these may include testing laboratories. These facilities aim to assist at various stages of the product value chain by providing infrastructure that is not easily available to individual enterprises.
Financial Assistance
Under this scheme, the Uttar Pradesh government provides financial assistance for the establishment of Common Facility Centers. The state government can provide financial assistance of up to 90% of the approved project cost for eligible projects, whose cost is up to ₹15 crore. However, the implementing SPV must bear at least 10% of the project cost itself. If the project cost exceeds ₹15 crore, the contribution of the state government will only be counted up to ₹15 crore.
The CFC scheme is designed to support groups of enterprises rather than individual enterprises. It attempts to strengthen district-level industries by preparing shared infrastructure through common technical facilities and services for a group of producers associated with an identified ODOP product.
This scheme also simplifies collaboration among artisans, manufacturers, producer groups, and MSMEs, giving them access to supporting infrastructure for production, quality improvement, and market readiness. By giving multiple enterprises the opportunity to use shared resources, this scheme seeks to strengthen the value chain of identified ODOP products and provide support to artisans, MSMEs, and producer groups associated with traditional industries.
