Advances in pretrained foundation models, synthetic data generation, and massive inference compute have created a compounding flywheel of improvement. The pace of acceleration at a triple exponential is incomprehensible. The trajectory of GenAI is such that the next decade could seemore technological progress than the last several decades combined. Large Language Models have led to Large Reasoning Models which are leading us into the Agentic AI and Physical AI Era. Not only are GenAI capabilities accelerating, but the costs of deploying these models are plummeting as efficiencies increase. This lowers the barrier to deploy AI, having large consequences upon industries.
We examine how GenAI is impacting three sectors: Robotics, Defense, and Biotech – and why forward-looking investors should take notice. In the coming weeks, we will publish a deeper dive on the candidates below and where we are investing.
The Plunging Cost of Intelligence
One of the most important enablers of AI’s spread is the rapid decline in computational costs. Key AI metrics like tokens generated per dollar and training cost are improving at staggering rates. These cost declines stem from multiple compounding factors: better chips, algorithmic optimizations, and smaller efficient models. Here are some current base rates on AI progress.
- Every 100 days the parameters a model needs to achieve the same capability decreases by half.
- Inference cost decreases by ~10x per year
- Training cost falls ~3x annually
These lower costs allow Gen AI to become increasingly multimodal (audio and visual), an evolution from Large Language Models.
Killer Robot Moment?
Historically, robots have been relegated to domain specific tasks such as automotive welding or home vacuums. This has led to successes such as iRobot which alone has shipped over 30mm units over the past 7 years. Although experiencing a modicum of success, the “Killer Robot” (we are not speaking about a Terminator here) still has yet to arrive. However, that is about to change.
A key challenge of robotics is their deterministic nature. The need for extremely high precision and feedback are very important (force and achieving millimeter level accuracy is hard). This need increases with the delicacy of the task. For instance, during the first iterations of LLMs, hallucinations were common and while harmless in text, hallucinations in real-world applications can be dangerous. New reinforcement learning methods pioneered by DeepSeek—without human feedback—allow GenAI to thrive in deterministic fields. General purpose models can be trained in physics simulations and use synthetic data to train for objective outcomes such as picking up objects or running. It essentially allows AI to “understand” the world. The era of synthetic data is here.
Physics simulations are critical for generating realistic synthetic data in robotics, allowing AI models to train and be tested in highly accurate virtual environments. These simulations enable billions of iterations, dramatically accelerating learning. This approach enabled AlphaZero to master Go and Chess through millions of self-play iterations. Examples of physics simulators used by AI today are NVIDIA Isaac (shown below) and Omniverse.
In a recent NVIDIA study by Jim Fan, it only took 2 hours of simulation for a humanoid to learn how to walk. And this was only a 1.5 million parameter model. An amazing achievement once thought to be computationally prohibitive. Engineers now train robots in simulation and transfer learning to physical systems, eliminating manual programming and costly real-world data. This significantly expands the capabilities of robots, enabling a broader range of autonomous tasks. To accelerate progress, NVIDIA has released GROOT, their Visual Language-Action (VLM) Model which is like ChatGPT for text. VLMs taken in pixels and outputs motor commands (instead of tokens). AI models will replace the PLC as the brains of the robot.
Robotics is entering a new era where AI-trained machines work safely alongside people. Leading adopters like Amazon have deployed over 750,000 mobile robots in warehouses worldwide. As the utility of robots increases and the costs to manufacture decrease, the form factors of robots will explode: Drones, Quadrupeds, and Humanoids are some of the ones existing today and will evolve into increasingly niche applications. In certain fields, as humanoid robots (or others) reach parity with human labor, entire industries will be upturned. The hardware costs of developing these robots has gone from $1mm in 2000 to about $50k today with room to fall significantly more. While humanoid production is still in its infancy with global production estimates currently below 10k units, we believe this number will asymptote very soon with 2026 production targets exceeding 100k units. Humanoids (and other form factors) will be the interface between moving bits to moving atoms.
Companies to watch:
- Harmonic Drive Systems (TK:6234) : Harmonic Drive Systems makes high-precision gears that are essential for accurate and reliable robotic movement. Its technology is used in fields like surgical robotics and semiconductor manufacturing, where precision is critical. As global demand for automation grows, Harmonic is positioned as a key supplier powering the next wave of robotics.
- Tesla (NYSE: TESLA ): Although there is some hype already, robots could be the real breakthrough where Tesla holds a lasting advantage. It is the only U.S. company building humanoid robots at scale, combining AI, advanced manufacturing, and full-stack integration. As automation needs rise, Tesla’s Optimus project could become a defining force in the future of labor. Tesla aims to produce 50-100k humanoids in 2026.
- BYD (HK:211) : BYD is often called the Chinese Tesla, with deep expertise in EVs, batteries, and vertical manufacturing. It is beginning to move into robotics and AI integration, positioning itself as a key player in China’s push toward automation. Although the robotics side is early, BYD has the scale, infrastructure, and government support to become a major force in robotics. BYD aims to produce 20k humanoid robots by 2026. Other players in China that are at the forefront of robotics are Unitree but it is a private company.
Drones and Digital Shields in Defense
Add rising geopolitical tensions to the rise of AI and you have a veritable cocktail of scenarios that can play out. The Ukraine Russia war has highlighted the changing landscape of war and the asymmetric opportunities that drones can have. A $500 dollar drone can now destroy a $10 million dollar tank, a 20k:1 ratio (drones have also taken down jets and ships). In 2024, Ukraine produced ~1.5 million drones and aims to triple this number in 2025. These drones were responsible for two-thirds of Russian losses, making them the most effective weapon in Ukraine’s arsenal. While these have been human operated, once AI enters the mix there will be a Cambrian explosion. Drone swarms that can operate in a centralized/decentralized fashion will be common and counter electronic equipment will have to evolve rapidly to keep up. Thanks to advances in edge processing chips and distilled models, drones will be able to navigate and communicate peer-to-peer without human control. The Pentagon has demonstrated swarms of over 100 micro-drones that were able to form an autonomous swarm exhibiting collective decision-making and adaptive behaviors.
Companies at the cutting edge of this include startups like Anduril, which has developed drones and is now building larger autonomous aircraft for the U.S. Air Force. Notably, Anduril and General Atomics were recently awarded prototype contracts for the Air Force’s Collaborative Combat Aircraft (CCA). Under CCA, the USAF is developing AI-enabled unmanned fighter jets designed to fly alongside human-piloted fighters in combat (shown below). The first two prototypes are set for flight tests this year, marking the first time the Air Force has given fighter aircraft designations to drones. This initiative signals that militaries view autonomous aircraft as the key to future air supremacy.
AI is reshaping the defense industry from drones in the skies of Ukraine to Pentagon procurement plans and presents a significant space race style investment landscape in the coming decade. Drones, Edge Processing, Cybersecurity will be the main drivers of defense spending.
Companies to watch:
- Parrot (PA:PARRO): PARRO was an early pioneer in consumer drones and achieved significant early success. However, driven by a challenging consumer market, PARRO has shifted its focus to enterprise and government contracts in Europe and the USA. PARRO is well-positioned to capitalize on increased defense spending particularly in Europe. Disclosure: We are long Parrot.
- Ambarella (NAS:AMBA) : Ambarella specializes in developing low-power, high-resolution video compression and computer vision processing semiconductors. It offers highly integrated, scalable chip solutions supported by comprehensive software development kits, enabling customers to efficiently create differentiated products. Its technology is used across a broad range of applications, including automotive cameras, security systems, and consumer devices, delivering advanced intelligent video capabilities. As the wave of robotics and edge AI devices emerges, the demand for low-power vision processing is expected to grow significantly. delivering system-on-chips capable of running computer vision and even language models on drones or smart missiles.
The End of Eroom’s law?
In biotechnology, generative AI is turbocharging the discovery of new drugs and therapies, compressing timelines from discovery to pre-clinical trials (We focus more on biotechnology, where transformation is just beginning, unlike medical scribe and radiology where the use cases that are further along). The landmark achievement in this space was DeepMind’s AlphaFold, which in 2021 solved the 50-year grand challenge of predicting protein structures from amino acid sequences- a previously intractable problem due to the 10^300 configurations a protein could become. AlphaFold achieved atomic-level accuracy in calculating protein 3D shapes, subsequently releasing an open database of over 200 million protein structures. This “protein universe” predicted by AI is a trove for discovering drugs. Any researcher can now look up the structure of any target protein in seconds (as simply as a Google search) instead of spending months or years crystallizing and imaging it.
The immediate impact of AI is on drug discovery, knowing a protein’s shape allows chemists to design molecules that “fit” into its pockets or modify its function. Biotech tasks often have measurable outcomes (e.g., Does this molecule inhibit this receptor? Does this protein fold correctly?) which then other parameters are optimized for such as solubility, stability. This makes it ideal for computational modeling or “in silico”. Like robotics, synthetic data will also prove useful here. Synthetic data is less of a problem in biotech than in other domains due to being grounded in science which makes feedback loop possible. Molecular simulations are also becoming very high fidelity as well, enabling more predictive in silico testing.
Traditional drug discovery has historically been a slow and expensive proposition with new drugs costing $2.5 Billion and taking more than a decade. Less than 10% of Phase I candidates receive FDA approval. Eroom’s Law, which states that there will be an ever increasing cost and declining efficiency of drug discovery and development over time might be coming to an end with AI. A recent example is Exscientia(a UK-based AI drug company) which designed a drug candidate for OCD that entered Phase I trials in a little over a year versus the ~5 years it typically takes.
A crucial enabler of AI will be lab automation. One can imagine a feedback loop – where AI models propose molecules/biologics, lab automation synthesizes and tests them, and the results are fed back to refine the model – allowing rapid iterations that home in on an optimal drug. In essence, it’s a form of optimization by AI that keeps improving suggestions based on real data. Recent case studies show up to 70 % fewer protein-production runs with the use of AlphaFold. Combined with recent FDA deregulation, the biotech setup looks very promising.
Companies to watch:
- Ginkgo Bioworks (NYSE:DNA): Ginkgo Bioworks is revolutionizing synthetic biology by integrating advanced automation and AI to streamline organism design. Their platform combines ultra-high-throughput screening, DNA synthesis, and metabolic engineering, enabling rapid prototyping and development of engineered organisms. Collaborations, such as their partnership with Google Cloud, further enhance their capabilities in protein modeling and AI-driven biological design
- Tecan (SW:TECN): Tecan is a leader in laboratory automation, offering AI-enhanced solutions like the Veya™ platform, which simplifies complex workflows and improves decision-making through real-time analytics . Their integration of AI-driven predictions with empirical wet-lab testing accelerates drug discovery by transforming inefficient processes into targeted workflows. Tecan’s end-to-end solutions support various applications, from single-cell dispensing to 3D spheroid analysis, enhancing reliability and throughput in research.
- Certara (NAS:CERT): Certara leverages AI and biosimulation to transform drug development, enhancing clinical trial design and execution through their platforms CODEX . It integrated advanced modeling and analytics, enabling life sciences organizations to streamline workflows and drive regulatory success . By applying generative AI to clinical trial documents, Certara accelerates data extraction and analysis, facilitating more efficient and effective drug development processes
Conclusion
Exponential improvements in AI will create nonlinear market demand. Small emerging markets can become enormous as capabilities cross certain thresholds. Generative AI is a fundamental acceleration engine for innovation. It will have rippling effects in robotics, defense, biotech and beyond. For investors, it means more complexity and staying abreast of new technologies, which companies and countries are leading, which are emerging, and how policies will affect the flow of capital and innovation. The importance of AI is well understood. What’s often overlooked, and what we aimed to demonstrate, is the magnitude of the quantum leaps underway in certain sectors.
INVESTMENT DISCLAIMERS & INVESTMENT RISKS
Past performance is not necessarily indicative of future results. All investments carry significant risk, and it’s important to note that we are not in the business of providing investment advice. All investment decisions of an individual remain the specific responsibility of that individual. There is no guarantee that our research, analysis, and forward-looking price targets will result in profits or that they will not result in a full loss or losses. All investors are advised to fully understand all risks associated with any kind of investing they choose to do.