Robot + AI: is it happening sooner than we thought?

As an investment theme, humanoid robots must be about the best suited to going viral via Instagram or TikTok. The demos floating around on social media seem to be more impressive each month.

So how much of this is hype backed up by smoke and mirrors - and how much is the real deal?

This week, we are taking a look at embodied AI and humanoid robots to find out.

What Happened In Markets This Week

Here’s a quick summary of what’s been going on:

🛢️ Oil supply shock pushes inventories to 8-year lows ( Reuters )

• What happened: Global oil inventories have dropped to around 101 days of demand, the lowest level in nearly eight years, with refined product stocks falling even faster to about 45 days. Even if normal supply via the Strait of Hormuz were to resume it could take months for inventory levels to normalize.
• How it impacts investors: Tight supply and rising geopolitical risk are driving oil prices higher, benefiting energy producers and related infrastructure while pressuring transport and consumer sectors. Sustained price spikes could increase recession risk and shift market sentiment broadly.
• Next steps: Learn more about the l ong-term effects of the Strait’s closure . Explore energy sector opportunities with your stock screener.

🤖 AI compute demand surges as Anthropic locks in massive capacity deal ( Yahoo Finance )

• What happened: Anthropic has agreed to rent the full capacity of SpaceX’s Colossus 1 data center, securing over 300 MW of power and more than 220,000 Nvidia GPUs. The deal adds to existing partnerships with major tech firms and comes as Anthropic explores future infrastructure, including potential orbiting data centers.
• How it impacts investors: The scale of this deal reinforces that AI infrastructure demand remains supply-constrained, supporting growth across chips, data centers, and power providers. It also signals continued capital intensity in AI, which may shape valuations and competition.
• Next steps: Dive into our Data Center (AI Picks and Shovels) Stocks collection to spot potential winners beyond the obvious names.

💾 Memory chip profits explode as AI demand drives historic boom ( WSJ )

• What happened : Samsung reported Q1 net profit above $30 billion, with around 94% coming from semiconductors, as memory prices surged nearly 100% quarter over quarter. Industry-wide, Samsung, SK Hynix , and Micron are expected to generate about $350 billion in net profit in 2026, with customers locking in long-term supply contracts and prepayments.
• How it impacts investors: The AI-driven memory boom is concentrating profits among chipmakers while creating supply constraints for downstream electronics. This could reshape margins across the tech sector and reinforce the dominance of semiconductor leaders.
• Next steps: Use the Stock Screener to identify semiconductor companies benefiting from AI demand and compare their fundamentals.

💊 Novo Nordisk gains as Wegovy pill drives record obesity drug demand ( Quartz )

• What happened: Novo Nordisk reported $354 million in Q1 revenue from its Wegovy pill, nearly double expectations, with prescriptions surpassing 2 million since launch. The company raised full-year guidance as both oral and injectable versions continue to grow, despite new competition entering the market.
• How it impacts investors: Strong early demand reinforces the scale of the GLP-1 market opportunity and intensifies competition between major players. Market share trends and sustained demand will be key drivers for valuation in the sector.
• Next steps: Review Novo Nordisk’s fundamentals and growth outlook with the company report .

Embodied AI: Structural Demand vs. an Autonomy Gap

In 2024 Goldman Sachs revised its global market forecast for humanoid robots upward by a factor of six, expecting US$38 billion in sales by 2035 . Morgan Stanley followed up a year ago, projecting a US$5 trillion global humanoid ecosystem by 2050 .

These figures contrast sharply with the current reality: fewer than 7,000 humanoid units were shipped to industrial buyers globally last year.

Nevertheless, investment is surging; China alone now hosts as many as 150 companies developing humanoids with state support . In January, Boston Dynamics unveiled the production version of its Atlas robot , and said production would begin immediately. Tesla’s Optimus is due to go into production in August, although prior targets have been cut considerably.

The shift to "Embodied AI"

Traditional industrial robots have been utilized for decades , typically as articulated arms performing single, repetitive tasks. Embodied AI represents a departure from these static machines, creating systems that can perceive, reason, and act within the physical world. This technology rests on four pillars :

1. 👀 Perception: Using LiDAR, cameras, and touch sensors to map the environment.
2. 🧠 Cognition: Utilizing LLMs and Vision-Language Models (VLMs) to process information.
3. 📝 Planning: Leveraging reinforcement learning to decide on action sequences.
4. 🕹️ Control: Converting decisions into physical movement through precision motors and actuators.

Humanoids, designed to replicate the human form to navigate human-centric environments, are the most prominent example of this.

Besides humanoids, embodied AI can take the form of:

• Autonomous vehicles , including robotaxis.
• Drones used for logistics, defense, law enforcement and agriculture.
• Other types of robots used for logistics, inspection, and handling hazardous materials. Boston Dynamics’ quadruped, Spot, is a commercially available example.

The economic driver: why now?

The surge in generalized robotics is driven by a convergence of technological capability and demographic necessity:

1. 🧑‍🏭 Labor Scarcity: In many developed nations (and others including China), more people are retiring than entering the workforce, creating a persistent labor shortage.
2. 📈 Return on Investment (ROI): Until recently, robots were too expensive and specialized to be cost-effective for many firms.
3. ⚙️ Generalization: Rapid advances in AI reasoning allow a single "device" to be trained for multiple tasks, working 24 hours a day. This multi-tasking capability is the "crucial piece of the puzzle" that makes the business case for humanoids viable in factories, homes and service industries,

Drones and other types of embodied AI have their place, but the widespread use cases for humanoids make the potential market so large.

Autonomy is still a work in progress

The demos may appear impressive, but the reality is that there’s a lot of ‘faking it till you make it’ going on. For both demonstrations and production, humanoids are often (if not usually) being remotely operated by humans.

As an example, 1X’s website states : “ NEO works autonomously by default. For any chore it doesn’t know, you can schedule a 1X Expert to guide it, helping NEO learn while getting the job done.” BTW NEO is available for pre-order ($20k to buy, $499/month to rent).

Morgan Stanley put it bluntly: "If a humanoid demo is not explicitly advertised as autonomous, one should assume it's tele-ops".

The hurdles

Judging by how far AI has come in the last few years, there’s no reason to think the autonomy gap won’t narrow over time. Whether it takes a year or 10 years is the $5 trillion question.

The brain (AI) is just one aspect of a reliable, autonomous robot. Humanoids have 30+ ‘joints’ moving independently, powered by 50+ motors. Walking and hand dexterity are still major challenges for human-like machines.

Beyond this complexity, there are some other major challenges and questions facing the industry:

• Can humanoids be hacked by cybercriminals?
• Can they be used to spy on their owners? Will they broadcast data back to the manufacturer?
• If a robot causes harm, who is liable?
• Will consumers be comfortable leaving humanoids in their homes all day?

There are also political and economic questions:

• Will humanoid workers face the same resistance as off-shoring and immigrant workers?
• Will robots be taxed to offset the impact on the labor market or government revenue?

Ultimately, policymakers may need to come up with a regulatory framework before widespread adoption occurs.

The middle path

The super-bullish outlook might be unrealistic in the near term, but there is also real demand for cost-effective automation, particularly for manufacturing and some service industries. Investors will need to make realistic assumptions, based on what’s actually possible, and the regulatory environment.

The main players in the industry

The humanoid race now includes quite a few companies, but only a few are publicly listed.

The largest are Tesla and Hyundai Motors, the majority shareholder of Boston Dynamics. In both cases, humanoids are unlikely to move the needle for quite some time, but they can still form part of the overall narrative.

This narrative outlines the case for Tesla as a Physical AI monopoly or platform .

The other listed companies pursuing humanoids include Xiaomi, Xpeng and Ubitech in China and Richtech Robotics in the US.

The legacy robotics and automation companies like ABB and Teradyne aren’t pursuing humanoids just yet. But they have long-term expertise in the space, and could enter later or form partnerships.

👉🏻 This watchlist includes both groups of companies:  Humanoid and Legacy Robotics Leaders

The US-China paradox

An interesting dynamic for this industry is the fact that most of the companies developing humanoids, and most of the component manufacturers are in China.

On the other hand, the leading AI models and chips required to power them are owned by US companies. This is som ething likely to play a role in the US/China relationship.

👉🏻 Learn more about why China is a  key player in emerging technologies .

The Robotics supply chan

For investors, the companies supplying the industry could be more interesting.

Elon Musk pointed out that Optimus includes over 10,000 unique components . Other robots in early R&D stages are estimated to embed as much as $50,000 worth of semiconductors . That will have to come down to reach the $20,000 price point manufacturers are aiming for, it’s still a lot of chips.

The components of a robot include:

• Both CPUs and GPUs which need to run models onboard.
• An operating system, AI models, and other software.
• Sensors and cameras.
• Batteries and chips to manage and distribute power. Efficient batteries are crucial to reduce the time spent charging rather than working.
• A unique array of motors, actuators and other components to rotate joints and move limbs.

Smaller companies could carve out lucrative niches developing specialized components within this ecosystem.

This watchlist includes the major players in this supply chain: Embodied AI: Picks and Shovels

💡 The Insight: Narratives over Numbers

With emerging industries like humanoid robotics, the quantitative data often fails to tell the whole story. You can’t simply extrapolate historical data into the future, and by the time the income statement reflects the story the ship has sailed.

F orecasts (like the $5 trillion TAM) are based on assumptions that can be rendered obsolete by a single technical failure or a regulatory shift.

As an example, analysts' forecasts for Ubitech Robotics are quite upbeat. But what catalysts are those numbers based on, and do you agree with them?

Because the data is noisy and the future is speculative, narratives become the essential tool for discipline. A strong narrative forces you to document

• your catalysts (what must go right for demand to scale?)
• your assumptions (units shipped, margins, etc)
• and your risks (what are policymakers doing?)

Doing this will give you conviction to pull the trigger when the price is right, and you’ll also know when reality isn’t playing out the way you thought it would.

Key Events Next Week

Monday

• 🇨🇳 China CPI YoY (April)
  • 📉 Forecast: 0.8%, Previous 1.0%
  • ➡️ Why it matters: Further easing may deepen deflation fears.

Tuesday

• 🇺🇸 US CPI YoY (April)
  • 📈 Forecast : 3.6%, Previous 3.3%
  • ➡️ Why it matters: This will be closely watched as energy prices have gone up.

Wednesday

• 🇺🇸 US PPI MoM (April)
  • 📉 Forecast : 0.3%, Previous 0.5%
  • ➡️ Why it matters: March PPI was driven by a 42% jump in diesel fuel; a cooling print would signal pipeline pressures easing.

Thursday

• 🇬🇧 UK GDP QoQ – Preliminary (Q1 2026)
  • 📈 Forecast: 0.3%, Previous 0.1%
  • ➡️ Why it matters: A rebound would ease recession fears, but energy shocks tilt risks toward a stagflationary outcome.
• 🇺🇸 US Retail Sales MoM (April)
  • 📉 Forecast: 0.1%, Previous 1.7%
  • ➡️ Why it matters: March's surge was driven by a 15.5% jump in gasoline station sales; a normalisation is expected.

Companies reporting the week include US tech names and prominent international companies:

• Cisco
• Applied Materials
• Circle Internet Group
• Nu Holdings
• Sea Ltd
• Barrick Mining
• Siemens
• Tencent
• Alibaba
• Sony Group

Have feedback on this article? Concerned about the content? Get in touch with us directly. Alternatively, email editorial-team@simplywallst.com

Simply Wall St analyst Richard Bowman and Simply Wall St have no position in any of the companies mentioned. This article is general in nature. We provide commentary based on historical data and analyst forecasts only using an unbiased methodology and our articles are not intended to be financial advice. It does not constitute a recommendation to buy or sell any stock and does not take account of your objectives, or your financial situation. We aim to bring you long-term focused analysis driven by fundamental data. Note that our analysis may not factor in the latest price-sensitive company announcements or qualitative material.