Headlines this week - Jan 5, 2025
A look at how capital is being deployed across future opportunities
This week in the future:
2025 could be “the year of the robots”, according to some forecasts. The use of generative AI to power them, and increasing interest from investors might accelerate commercialization and practical applications. But there are some challenges ahead (some of them linked to “Moravec’s Paradox”)
Nvidia is trying to capture the opportunity, developing hardware and software for AI-powered robots, investing in robotics companies, and providing tools for robot development.
This is part of a more general effort by Nvidia to diversify, and become a more general enabler of AI solutions. Within this strategy, they just announced they're open sourcing a tool previously working only with Nvidia GPUs, that now others could use to manage different chips too.
Tesla is losing market share in electric vehicles. They just reported their first decline in deliveries since 2011. Meanwhile, Chinese vendors keep accelerating (with BYD at the forefront)
But investors no longer see Tesla as an electric car vendor. So there was no significant reaction to the company’s negative news. Today Tesla is perceived as an AI player, with exposure to robotics and autonomous vehicles
The AI investment tsunami is expected to continue. Microsoft just announced plans to spend $80bn (close to 30% of the company’s annualized revenues) on AI data centers
Investors are excited about biotechnologies aiming at expanding human lifespan (“Longevity Tech”). Money is flowing to startups working in the space. But massive uncertainties remain, from technical challenges and commercial barriers to regulatory and ethical issues
Nuclear energy is becoming a key tool to accelerate the “green energy transition”.A $1bn deal of Constellation Energy (a leading nuclear energy company in the US) with the US government illustrates the trend. And more deals like this are expected in the coming months.
Military applications keep pushing for a new “space age”. Strategic considerations for defense systems might benefit Blue Origin as an alternative to SpaceX, with buyers trying to reduce their dependence on a single vendor.
Will we live in “permanent inequality” after the singularity? Scott Alexander discusses scenarios for a post-”AI singularity” future. The good news are that this future might still largely depend on us, with scenarios of extreme inequality equally possible as others where the benefits of AI are more widely distributed
Happy new year. Please share this with your friends if you think it is interesting. And give me feedback if you’re missing something :-)
1 - Population & natural resources
Water shortages emerging as a potential threat for the EU. EU Environment Commissioner Jessika Roswall has just stressed the urgency of addressing water shortages, which are impacting businesses, agriculture and other crucial sector. She’s claiming for large investments in infrastructure to deal with this problem. EU’s energy fixation has left water crisis unchecked, top official warns
Underwater habitats: a new opportunity? Deep, a startup, is working to make it possible for scientists to live and work underwater for weeks, months, or even years, significantly expanding the scope of marine research and potentially opening up new possibilities for human habitation. The company’s hope is to achieve a permanent human presence in the ocean by 2030. In 2025, People Will Try Living in This Underwater Habitat
Biotech
Researchers are looking to create artificial biological life in the lab. MiniLife, a European research initiative, aims to create simple forms of life from scratch in the lab. They don’t rely on existing biological building blocks, but rather start from simple chemicals, looking to trigger autocatalytic reactions and self-replication. The objective is to better understand the origin of life and, potentially, to help develop new drugs and new materials. Scientists aim for ‘Darwinian evolution’ with artificial life project
Longevity / Health
Longevity technologies: significant (high profile) investments, but still plenty of uncertainty. Longevity startups raised $430 million in venture capital in the first nine months of the year, and one company, BioAge Labs, raised almost $200 million when it went public. Prominent investors, like P. Thiel, R. Kurzweil, J. Bezos and Y. Milner, are involved. But massive uncertainties remain, from technical challenges and commercial barriers to regulatory and ethical issues. Do not bank on living forever in this lifetime
Uncertainties are also turning healthcare (in general) into a challenging space for startups. The emergence of large tech companies (like Amazon) as competitors and the emphasis on scale and proprietary data are critical factors shaping the industry. The potential of AI in drug discovery, diagnostics, and healthcare productivity enhancement is considerable, but it also comes with important challenges that need to be addressed. Why returns have been slow to follow investment in digital health
A typical challenge is collateral damage to other industries. E.g. food / sugar and weight-loss drugs. So-called “GLP-1 drugs” curb users’ appetites, and this makes them effective for tackling obesity and related conditions. But their effects could lead to a reduction in sugar consumption, potentially impacting the sugar market, and this is causing concerns in the industry. For example, stocks of Mondelez and PepsiCo took a hit after Walmart reported that customers using these drugs were buying fewer groceries, and Hershey has also noted a mild impact on sales. Will weight-loss drugs lead to upheaval in the sugar market?
A new wave of personalized medicine could emerge with the discovery of somatic mutations. Somatic mutations, genetic changes that occur after conception and are not inherited from parents, have now been shown to lead to conditions that were not previously considered genetic. This is expected to drive new developments in personalized medicine, improved diagnostic approaches, targeted therapies, and a deeper understanding of complex diseases. Doctors Thought They Knew What a Genetic Disease Is. They Were Wrong.
Space
A golden age for space technologies is expected, driven by the needs of the defense industry. Increased investment is expected in the space sector from both large defense contractors and private equity/venture capital firms, which expect to capture the opportunity of military space systems (e.g. space communications). Defence industry set for deal surge as companies look to expand in AI and space
Can Jeff Bezos’ Blue Origin help governments diversify space suppliers? In the US, Elon Musk’s closeness to the new US government may give SpaceX an additional advantage. But in other countries, this could turn into an opportunity for Blue Origin, which is now trying to catch up. Elon Musk is standing in the way as Jeff Bezos reaches for orbit
Materials
Startups using AI to optimize mining efficiency are raising massive investments. KoBold Metals, a mining start-up backed by Bill Gates and Jeff Bezos, is using AI to identify critical mineral deposits needed for the energy transition and has raised nearly $3 billion in funding, positioning itself to become a major player in the mining industry. Mining start-up backed by Bill Gates and Jeff Bezos valued at $2.96bn
2 - Efficiency & Productivity
Is the US on the way to solve the “productivity paradox”? Recent data suggests this may be the case, due to an increasing adoption of new technologies (including AI), an improvement business operations, and a shift of workers to higher-skilled jobs. However, it remains to be seen if this will be sustainable (but innovation in the “atom world” could contribute to that…) The American Worker Is Becoming More Productive
Science is apparently making progress. But Physics no so much… In the current debate about the potential stagnation of science, the FT just published a summary of progress this year, which suggests that scientific progress might actually be accelerating. Breakthroughs in cosmic exploration (via the James Webb telescope), human health, biotechnology and robots are explicitly discussed. Physics is interestingly (almost) absent… A year of scientific breakthroughs
Energy
Nuclear
Growth in nuclear becoming key for the energy transition. Constellation Energy, the company with the largest nuclear energy production footprint in the US, has secured a $1 billion deal with the US government to provide electricity and services, which will fund upgrades to their current infrastructure. This confirms the critical role that nuclear is expected to play as a source of clean energy, in the coming years. Constellation Energy Signs $1 Billion Power Deal With US Government
Renewables
Local opposition to infrastructure deployments remains a key challenge. The case of wind farms in the UK, where local communities are opposing deployments (in a classical “Not In My BackYard” behavior) clearly illustrate the point. As nuclear deployments start to proliferate again, we should probably expect more of this. The looming battle over Britain’s wind farms
New Transport Technologies
Electric Vehicles
Tesla is feeling the competitive pressure from Chinese rivals. Tesla's annual deliveries declined for the first time since 2011, despite a record number of sales in the fourth quarter, signaling increased pressure from Chinese rivals. Tesla’s annual deliveries drop for first time since 2011
Tesla’s market share in the US is slipping, amid increased competition and a cooling consumer interest in its older models. The company is planning to launch a new, lower-cost model in the first half of 2025, to respond to this trend. Tesla’s Global Vehicle Deliveries Fell in 2024 for the First Time in Years
Investors’s reaction to this shows how they now perceive Tesla as an AI company, rather than an electric car vendor. Tesla has been shifting its focus to autonomous driving, AI, and robotics as its core auto business faces challenges, and valuation is primarily driven by its potential in AI and robotics rather than its current auto business. So the company's declining car sales in 2024 are overshadowed by investor optimism regarding its future technologies. Tesla’s Car Business Isn’t Really in the Driver’s Seat
Meanwhile, in China, BYD and Xiaomi exceeded their sales targets. BYD sold a record 4.3m EVs and hybrids in 2024, and nearly matched Tesla in pure EV sales, with 1.76m units. The Chinese EV market is currently under intense competition and price wars, and this is leading to consolidation, with some start-ups collapsing, while others, like Geely, are streamlining operations. China’s electric-vehicle leader BYD posts record sales in 2024
in Europe, the UK is competing with Germany to be the continent’s largest EV market, The UK saw a record 382,000 EVs sold in 2024 (vs. almost 350,000 sold in Germany). However, the UK is still below its government targets, and the EV market is dependent on discounts and incentives to spur sales. UK vies with Germany to be European EV champion
Supersonic Airplanes
China wants to dominate the global aircraft market. China is increasing its efforts to challenge the dominance of Boeing and Airbus in the aircraft market, with its state-owned manufacturer Comac seeking international certifications for its C919 jet and developing its widebody C929, which is part of a broader strategy to move up the technology value chain and compete with Western aerospace manufacturers. China steps up drive to break Boeing and Airbus grip on plane market
Artificial Intelligence
Nvidia might be on the way to become an AI “generalist”. Nvidia is strategically investing in numerous AI startups, including some of its own customers, and acquiring AI companies to expand its ecosystem and ensure demand for its chips, while potentially shifting from primarily selling chips to becoming a key enabler and backer of the broader AI industry. Robotics (see below) is one of the opportunities that the company is now actively exploring. Nvidia invested $1bn in AI deals in 2024
Consistently with this, Nvidia is open-sourcing a piece of software initially designed to support only Nvidia GPUs. This suggests a strategic effort to broaden its adoption and empower the wider AI ecosystem. By enabling the software to extend its reach beyond its own hardware, Nvidia could be positioning itself as a more indispensable technology partner and thus expanding its reach into the AI industry beyond just hardware. Nvidia to open-source Run:ai, the software it acquired for $700M to help companies manage GPUs for AI
AI: Apps
B2C
YouTube expects AI to catalyze revenue growth. The company is already generating $50bn / year for Google, and they expect additional growth to be catalyzed by the elimination of barriers to content creation that is coming through the use of Generative AI tools. So we should prepare for much more AI-generated video content in the app i’m the coming months / years. YouTube chief Neal Mohan bets on AI and ‘creators’ to supercharge growth
B2B
Enterprise deployments of AI agents will require new software solutions. Some analysts expect an acceleration of generative AI deployments (in the shape of agents) in business organizations during 2025. For this to happen, companies will need the infrastructure to manage multiple AI agents and applications (already described as “AI orchestration”). Why 2025 will be the year of AI orchestration
AI: Robots
More comments about an imminent robotics revolution, driven by AI. Robots are moving beyond controlled factory settings and entering public spaces. This shift is accelerated by generative AI, and substantial investments, so insiders expect 2025 to be a “turning point” year. However, robots may not be ready to excel in all human interactions without some stumbles, as the Moravec paradox is still pretty much applicable (things looking trivial for humans are often very difficult for robots). AI Robots Are Entering the Public World—With Mixed Results
Challenges still remain in making robots operational “in the wild”. Even if AI advances are definitely a catalyst, scaling production, reducing costs, and enhancing the accuracy of robot products are still pending issues. AI does not mean the robots are coming
Plenty of things could go wrong in this initial phase, according to the FT’s Lex Column. Initial adoption is still limited by practical problems like high maintenance costs, need for frequent charging. Also, the need of bipedal (“humanoid”) designs for many applications is still an open question. March of the humanoids still has some way to go
This is one of the fields where Nvidia is looking for diversification. The company is strategically positioning itself as a key player in robotics, by developing hardware and software for AI-powered robots, investing in robotics companies, and providing tools for robot development. Nvidia bets on robotics to drive future growth
Military applications still look like a “low hanging fruit” for robots. Ukraine is moving fast with this, enhancing drone automation for front-line fighting with a goal of increasing the effectiveness of drone strikes. Ukraine Advances Killer Robot Drones With More Automation, Efficiency
AI: Security & Safety
AI chatbots open new spaces for privacy regulation. The recent settlement of a case against Apple related to users’ conversations with Siri, could be just an initial example. Apple Agrees to Pay $95 Million to Settle Lawsuit Over Recordings Tied to Siri
AI is making phishing attacks more effective. AI-generated phishing scams are becoming more sophisticated and personalized, making them harder to detect. AI bots can quickly analyze large amounts of data, such as a company's tone and style or an individual's online presence and social media activity, to create convincing and personalized phishing scams. Cybersecurity experts have a new challenge… AI-generated phishing scams target corporate executives
AI: Infrastructure
The AI investment tsunami is expected to continue. Microsoft just announced plans to spend $80bn (close to 30% of the company’s annualized revenues) on AI data centers. The tech industry is now officially a (highly) capital intensive one. Microsoft to Spend $80 Billion on AI Data Centers This Year
AI infrastructure deployments are accelerating in Asia. Macquarie anticipates a rise in infrastructure deals in the Asia-Pacific region in 2025, particularly in digital infrastructure and renewable energy, despite economic challenges. Asia infrastructure to drive more deals after $15bn data centre sale, Macquarie says
AI: Chips
The semiconductor industry is changing rapidly, and Intel is a victim of that. First, the semiconductor industry is moving away from traditional x86 architectures towards ARM-based designs, driven by the need for customization, speed, and efficiency. As a consequence, Intel, once the leader in CPU production, is losing ground to competitors like AMD in the data center market. Furthermore, the increasing demand for GPUs, crucial for AI applications, has put Intel at a disadvantage. Intel is struggling to address these problems, in part due to its vertically integrated model. Intel’s Problems Are Even Worse Than You’ve Heard
3 - Economic / Business trends
Tech & Geopolitics
China wants to protect its electric car industry from disruptions in the global chip supply chain. The Chinese government wants to force the local electric car industry to use of China-made chips. This protects electric car vendors from external disruptions in the chip supply chain, but it also could help reinforce the local chip production (a geostrategic objective for China right now). Beijing Pushes to Use China-Made Chips in Its EVs
New rules against investment in Chinese technology make life more difficult for US investors. US-based VC funds will now have to look for ways to certify that their money is not being used to develop AI, quantum computing, or other advanced technologies that American policymakers view as “strategic”. US investors in China venture funds brace for new tech rules
France might be Europe’s (only?) hope to become a global AI player. France has the ambition and the government support, it also has the (high-skilled) talent required for this, and (according to the author) the entrepreneurial spirit. But political instability and Europe’s regulatory approach (again!) are significant challenges. Can France become a global AI powerhouse?
Business trends
Investors’ money is concentrating on a fewer number of Venture Capital funds. The number of active venture capital investors has dropped more than a quarter from a peak in 2021 (from 8,315 to 6,175). Number of US venture capital firms falls as cash flows to tech’s top investors
Crisis at Intel and Boeing: why focus on short-term profits may be dangerous in the long term. According to this WSJ analysis, Intel and Boeing suffered from a similar strategic management pitfall: a focus on short-term profitability at the expense of long-term innovation and quality. This led to a decline in their market position, which even technically-minded leaders struggled to reverse. The authors present this as something to learn by the rest of us, but practical applications are not so clear… Intel’s Grim Lesson for Boeing: Sometimes Mr. Fix-It Is Too Late
The World in the future
The FT summarizes key global trends for 2025, and many of them are directly related to technology: The author mentions the emergence of “Agentic AI”, the global commercial wars in semiconductors, the turmoil expected for the electric car industry (with demand cooling down except in China), and (at least initially) a good year expected for cryptocurrencies. Forecasting the world in 2025
Will we have “permanent inequality” after the singularity? Scott Alexander discusses scenarios for a post-singularity future. The baseline is based on a thesis by L Rudolf L, an effective altruist who argues that after the point when AI is able to assume most of our jobs, a "technofeudalism" status quo will emerge, involving a small group of super-rich individuals controlling most of the wealth and resources, while the majority of people live with a much lower standard of living. Scott explores several reasons why this "technofeudalist" future might not happen, and paints a picture of a future that is largely defined by the development of AI, but is not predetermined. The future could be one of extreme inequality, or one where the benefits of AI are more widely distributed. It's Still Easier To Imagine The End Of The World Than The End Of Capitalism