Jake Loosararian: Robotics must prioritize data collection for efficiency, the impact of Nvidia’s dominance on hardware diversity, and the crucial role of determinism in future advancements | TWIST

Key takeaways

• Robotics should prioritize data collection to optimize performance and decision-making.
• Industries like energy and defense are increasingly leveraging robotics for operational efficiency.
• The future of robotics is promising, but safety and reliability through determinism are crucial.
• Consolidation around Nvidia limits hardware diversity, impacting AI development.
• Robotics can enhance efficiency in industries with high energy costs and frequent shutdowns.
• GPUs have become vital for scaling AI applications, especially in chat-based models.
• Fragmentation in hardware compatibility is due to proprietary software systems.
• CUDA is outdated for modern systems, indicating a need for updated GPU software.
• Heterogeneous systems enhance computing flexibility and scalability.
• Enterprises seek hardware flexibility to avoid vendor lock-in.
• The pragmatic impact of AI and robotics is a focus for sectors like energy and defense.
• Determinism in robotics ensures safety and reliability in AI applications.
• The rise of chat-based models has driven GPU importance in AI.

Guest intro

Jake Loosararian is the CEO and co-founder of Gecko Robotics, a company deploying purpose-built robots and AI for mission-critical infrastructure inspection across energy, defense, and manufacturing. In 2012 as a student at Grove City College, he built his first wall-climbing robot in a dorm room to solve persistent downtime at a local power plant, launching the company in 2013. Gecko now manages over 500,000 critical assets for Fortune 100 partners and the US Air Force and Navy, reaching unicorn status with a $1.25 billion valuation in June 2025.

The role of data in robotics

• The idea of gathering information and data using robotics to help drive better outcomes

  — Jake Loosararian

• Robots should not be built just for the sake of building; they must serve a purpose in data collection.
• Data-driven robotics can prevent a commoditized future in the industry.

• If you’re building robots just to build robots… it leads to a commoditized future

  — Jake Loosararian

• Understanding the role of data is crucial for optimizing infrastructure performance.
• Robotics in infrastructure is about improving decision-making through data.

• The pragmatic impact of artificial intelligence… can potentially drive better decisions

  — Jake Loosararian

• Data collection is essential for enhancing operational efficiency in critical sectors.

Robotics in energy and defense

• Energy, oil, gas, and defense sectors focus on the pragmatic impact of robotics.

• The energy, oil and gas companies… are completely looking at how impactful can robotics be

  — Jake Loosararian

• Robotics and AI integration is enhancing operational efficiency in these industries.
• The defense sector is exploring robotics for improved decision-making.

• The department of war are completely looking at how impactful can robotics be

  — Jake Loosararian

• Robotics helps address challenges in industries with high energy costs.

• Robotics can significantly improve operational efficiency in industries facing high energy costs

  — Jake Loosararian

• The focus is on how robotics can drive better outcomes in energy and defense.

Future of robotics and determinism

• The future of robotics is optimistic but requires a focus on determinism.

• I am very excited and optimistic about… what the future will be with robotics

  — Jake Loosararian

• Determinism ensures safety and reliability in robotics applications.

• The key is being deterministic… that’s maybe where we’re lacking a bit

  — Jake Loosararian

• Safety and reliability are critical in the rapidly evolving field of robotics.
• Determinism balances innovation and safety in robotics.
• The focus on determinism addresses potential safety concerns in AI.
• Ensuring reliability in robotics is crucial for future advancements.

Hardware diversity and Nvidia’s dominance

• The consolidation around Nvidia limits hardware diversity in AI development.

• A lot of the world is really consolidated around the Nvidia platform

  — Jake Loosararian

• There is a need for more hardware vendors to foster innovation in AI.

• We want more hardware vendors in the space

  — Jake Loosararian

• Nvidia’s dominance impacts the diversity of AI hardware options.
• Hardware diversity is crucial for fostering innovation in AI.
• The current landscape of AI hardware needs more competition.
• Consolidation limits the potential for diverse AI hardware solutions.

The importance of GPUs in AI

• GPUs have become essential for scaling AI applications.

• GPUs have captured the world… the inference side of it is huge

  — Jake Loosararian

• The rise of chat-based models has driven the importance of GPUs.
• GPUs enhance computational capabilities in AI technologies.
• The role of GPUs is critical for inference tasks in AI.
• The evolution of AI technologies has increased the demand for GPUs.
• GPUs are vital for enhancing AI computational power.
• The importance of GPUs in AI continues to grow with technological advancements.

Fragmentation in hardware compatibility

• Fragmentation arises from the lack of a unifying software layer.

• Hardware companies don’t get along… they build software for their chips

  — Jake Loosararian

• Proprietary systems contribute to hardware compatibility issues.
• The competitive dynamics between hardware companies lead to fragmentation.
• Proprietary software solutions impact industry fragmentation.
• Compatibility issues arise from the lack of a unified approach.
• The impact of proprietary software on hardware systems is significant.
• Fragmentation affects the overall efficiency of hardware systems.

The need for updated GPU software

• CUDA is outdated for modern systems and generative AI.

• CUDA… is the shining star of system software for GPUs but it’s 20 years old

  — Jake Loosararian

• There is a need for innovation in GPU software for current technology trends.
• Existing GPU software may not meet the requirements of modern advancements.
• The relevance of CUDA is questioned in the context of new technologies.
• Modern systems require updated GPU software solutions.
• The evolution of technology demands innovation in GPU software.
• The need for updated software is critical for advancing AI capabilities.

Heterogeneous systems in computing

• Heterogeneous systems enhance flexibility and scalability in computing.

• You get these heterogeneous systems where you have different architectures

  — Jake Loosararian

• Different hardware architectures communicating enhances computing capabilities.
• Heterogeneous systems are vital for modern computing architecture.
• The impact of heterogeneous systems on enterprise flexibility is significant.
• Enterprises benefit from the flexibility offered by heterogeneous systems.
• The shift in computing architecture influences technology investments.
• Heterogeneous systems play a key role in future computing developments.

Avoiding vendor lock-in with hardware choices

• Enterprises desire the ability to choose between different hardware systems.

• It gives enterprises choice… they want choice to be able to adopt other systems

  — Jake Loosararian

• Avoiding vendor lock-in is a critical concern for enterprises.
• Flexibility in technology choices is essential for enterprises.
• Enterprises seek to avoid dependency on a single hardware vendor.
• The ability to choose different systems enhances enterprise flexibility.
• Vendor lock-in poses challenges for technology adoption.
• Enterprises prioritize flexibility in hardware choices to enhance innovation.