Non classé

The AI Wars: Battlefronts, Breakthroughs, and the New Era of the Industrial AI (R)Evolution

Published

2 ans ago

16 janvier 2025

The Ai Wars: Battlefronts, Breakthroughs, And The New Era Of The Industrial Ai (r)evolution

Collin Masson, Director of Research at ARC Advisory Group. Colin heads up ARC’s research into Industrial AI.

All supply chain vendors seek to position themselves as leaders in supply chain AI. But there is a larger AI ecosystem. Emerging leaders understand the AI ecosystem and have the right partnerships. The current AI landscape can be viewed as a series of “wars,” where companies and organizations are battling for dominance across various technological and market “battlefronts”.

This analogy is not just a matter of abstract concepts; it is about real-world investments, strategic partnerships, and the tangible products being developed that are shaping the future of industrial AI. Let’s revisit the key battlefronts I identified in the AI Wars and examine the flurry of AI announcements in 2024 for proof that this analogy is useful for contextualizing the chaos and the real dynamics at play in the industrial AI arena.

Datacenter Hardware: The demand for powerful computing to train ever larger and more accurate AI models is insatiable. The battle here is to develop hardware that can handle this massive computational load efficiently and cost-effectively.

The competition in this space is intense, as evidenced by the recent announcements from multiple major players. Nvidia continues to dominate with its high-performance GPUs, but companies like AMD and Intel are rapidly developing their own competitive offerings.
AMD unveiled an expanded roadmap for its Instinct accelerators, with the MI325X slated for late 2024 and the MI350 series promising a 35x increase in AI inference performance by 2025.
Intel has introduced its Xeon 6 processors for servers, aiming to offer competitive performance for AI workloads.
AWS, Google, and Microsoft are also investing heavily in custom AI chips to reduce their dependence on NVIDIA and optimize performance and cost.
AWS has custom AI chips—Trainium and Inferentia, for training and running large AI models. AWS has also embraced Nvidia’s H100 GPUs as part of Amazon’s EC2 P5 instances for deep learning and high-performance computing. AWS also announced new Amazon EC2 P5en instances with Nvidia H200 Tensor Core GPUs and EFAv3 networking.
Microsoft is leveraging its Azure Maia AI Accelerator optimized for AI and generative AI, as well as its Azure Cobalt CPU, an Arm-based processor designed to run general-purpose compute workloads on the Microsoft Cloud. Microsoft has also integrated NVIDIA’s new Blackwell (H200) chip and AMD’s ND MI300X V5 into its Azure supercomputing infrastructure.
Google has developed multiple generations of its Tensor Processing Units (TPUs), which are custom-built ASICs optimized for TensorFlow and used by Google Cloud for machine learning workloads. Google is also reportedly working on its own Arm-based chips. Additionally, Google has announced the general availability of its sixth-generation Trillium TPU, which they used to train Gemini 2.0.

These moves highlight the fierce competition to provide the infrastructure necessary for continued AI innovation and scale adoption, in the very active datacenter hardware battlefront.

Edge Hardware: The battle for edge hardware also intensified in 2024, as companies sought to deploy AI capabilities closer to the source of data. The focus is on creating AI-optimized chips and hardware for edge devices, making AI more accessible and practical for a wider range of applications.

Google’s Edge TPU is a purpose-built ASIC designed to run AI at the edge with high performance in a small and energy-efficient footprint. In addition, Google’s Pixel phones are equipped with a Tensor G3 chip, an AI powerhouse capable of 38 TOPS.
Apple Intelligence demonstrates a clear push for on-device AI processing, with new AI-driven tools enhancing productivity across their operating systems, with a heavy emphasis on privacy and Edge AI. This puts pressure on other device manufacturers to follow suit.
Microsoft’s Copilot+PCs represent a big bet on edge AI, with new silicon capable of 40+ TOPS and prioritizing power efficiency. This initiative is bringing powerful AI capabilities directly to user devices, with the first wave of Copilot+ PCs coming from Microsoft Surface and OEM partners such as Acer, ASUS, Dell, HP, Lenovo, and Samsung.
Qualcomm announced its latest Edge AI Box solutions, further demonstrating the expansion of AI capabilities at the edge. Qualcomm’s Edge AI solutions use Snapdragon X Elite chips, which are capable of 45 TOPS.
Nvidia’s Jetson Orin Nano Super Developer Kit is a new compact generative AI supercomputer that is designed to provide increased performance at a lower price. By providing a powerful yet accessible platform, the Jetson Orin Nano enables developers and researchers to innovate in edge AI. The ability to run AI models directly on devices without a constant cloud connection is crucial for applications requiring real-time responses, such as industrial automation, robotics, and autonomous vehicles.

These developments underline the importance of edge computing as a perhaps the most important battleground for the industrial sector in the AI Wars, where companies are competing to bring AI capabilities closer to the source of data, their factories, distribution networks and grids, and their customers.

General Purpose AI Software Platforms: Modernizing the Technology Stack for AI

The competition to deliver comprehensive AI software platforms escalated considerably in 2024. The goal of these platforms is to provide a versatile set of tools for training, validating, and deploying AI models across a wide range of use cases. The battle for general purpose AI software platforms is intense with all major cloud providers offering a variety of tools and platforms.

In late 2022, OpenAI arguably ignited the “AI Wars” with the release of ChatGPT 3.5, which brought a new level of accessibility and capability to generative AI. This event marked a turning point, moving AI from a primarily research-focused area into the mainstream consciousness, triggering a “mass scramble among businesses trying to implement the latest advances in generative artificial intelligence”. This also caused a surge in investments into AI startups, as evidenced by the fact that the companies on the 2024 AI 50 list have raised a total of $34.7 billion in funding.

OpenAI’s “12 Days of OpenAI” event showcased its continued efforts to enhance its competitive position in the AI market. The announcements demonstrate that OpenAI is actively refining its offerings to gain a larger share of the broader AI market, which is experiencing rapid growth across industries. Key announcements from the event include:

Introduction of ChatGPT Pro: This broadened the usage of frontier AI.
Updated OpenAI o1 System Card: This highlighted safety improvements, robustness evaluations, and red teaming insights.
Realtime API Improvements and a New Fine-tuning Method: These enhancements will assist developers in building more effective and efficient AI applications.
New Tools for Developers and OpenAI o1: These appear to be aimed at helping developers create and deploy AI solutions more easily.
ChatGPT Search: This feature gives users a way to get answers from relevant web sources.

By focusing on developer tools, improving model safety and performance, and expanding the functionality of ChatGPT, OpenAI is taking significant steps to maintain its position and compete with new LLMs.

Microsoft is significantly expanding its Azure AI capabilities with new tools such as the Azure AI Foundry SDK and portal, enabling developers to customize, test, deploy, and manage AI apps and agents with enterprise-grade control. The company is also introducing the Azure AI Agent Service to enable professional developers to orchestrate, deploy, and scale enterprise-ready agents. Also, a strategic alliance between C3 AI and Microsoft will make C3 AI’s enterprise AI software available on the Microsoft Commercial Cloud portal. For additional ARC insights read “Microsoft Ignite 2024: Key AI Announcements for Industrial Organizations”.
AWS continues to expand the capabilities of Amazon Bedrock, offering new features to help businesses build faster, more cost-efficient, and highly accurate models. AWS is also expanding its range of AI services and making them easier to use. For additional ARC insights read “AWS re:Invent 2024 Prepares Developers for AI at Scale in 2025”.
Google’s latest AI announcements include the release of Gemini 2.0, its most capable multimodal AI model, and new state-of-the-art video and image generation models, Veo 2 and Imagen 3, available on Vertex AI. Google is also introducing Agent Workspaces, bringing AI agents and AI-powered search to enterprises. These advancements are aimed at improving productivity, automating processes, and modernizing customer experiences through the use of AI agents.

These announcements demonstrate a clear battle for mind and market share, with each company striving to provide the most comprehensive and user-friendly AI platform for startups, ISVs, and enterprise developers.

Edge AI Software

For many industries, and AI use cases, it is a hybrid world that needs some training and lots of inference to happen on edge devices. Therefore, for scale adoption of AI, many of those leading AI research and development are focusing on reducing the complexity and cost of deploying AI models to edge devices.

NVIDIA is advancing physical AI with accelerated robotics simulation on AWS, showcasing its focus on edge AI in robotics. Field AI is building robot brains that allow robots to autonomously manage industrial processes, and Vention creates pretrained skills to ease development of robotic tasks, both showcasing NVIDIA and AWS platforms. NVIDIA’s 2024 edge AI software announcements focus on making AI more accessible and practical for robotics and industrial applications. By developing platforms such as Isaac Sim and Jetson, providing pre-trained skills for robots, and introducing microservices for multilingual AI, NVIDIA is facilitating the deployment of AI at the edge. These developments help enable real-time data processing, reduce the reliance on cloud connectivity, and democratize access to advanced AI technologies in industrial and robotic contexts.
Microsoft is also focusing on edge devices with the Windows Copilot Runtime APIs, which brings on-device machine learning to enterprise apps. The company’s acquisition of Fungible, a company that develops data processing units (DPUs) optimized for AI workloads, is another key aspect of its edge AI hardware strategy. Microsoft plans to use Fungible’s DPUs to accelerate the performance of Azure IoT Edge and other edge AI solutions.
Qualcomm announced its latest Edge AI Box solutions, which represent the cutting edge in security and surveillance space. Qualcomm’s Edge AI Box solutions help upgrade existing camera and security assets into smart IoT- and 5G-supported networks. The company’s solutions are designed to modernize older systems, bringing them up to date with the latest AI and networking technologies.

These developments highlight the push for edge AI in a variety of applications, from robotics to security, with companies working to make AI more accessible and practical on edge devices.

Data and AI Model Marketplaces and Exchanges

These platforms are becoming critical battlegrounds where companies compete for data and pre-trained AI models.

The emergence of Data and AI Model Marketplaces and Exchanges is a significant battlefront in the AI Wars, as companies are realizing the importance of data for training AI models.
The Microsoft Azure AI model catalog is where various industry-specific AI models are made available by companies like Bayer, Cerence, Rockwell Automation, Saifr, Siemens, and Sight Machine. These models are pre-trained with industry-specific data to address a customer’s top use cases.
Amazon Bedrock Marketplace allows access to various AI models and tools, providing a venue for companies to find the right resources to build their AI capabilities.
Microsoft Fabric is designed to allow any app or data provider to bring data into OneLake. This is where data providers can directly write change data into a Mirrored Database in Fabric, which demonstrates the battle for data control and dominance.

These marketplaces are not just about selling AI models, but also about the control of training data and data sovereignty, with companies and nations vying for control over their data.

AI Startups: The Guerilla Innovators in the AI Wars

At the forefront of the competition are innovative AI startups reshaping established markets with groundbreaking solutions. These startups serve as “guerrilla innovators,” propelling advancements in industrial automation, software, and processes through AI, computer vision, and robotics. Unconstrained by legacy systems, they can swiftly adapt and deliver transformative technologies to the market.

Focus on Specific Industrial Needs: While many AI startups are focused on general-purpose AI solutions, others are targeting specific niches within the industrial sector, demonstrating the versatility and broad applicability of AI technology. A small sample of startups in the industrial sector include:

Anduril Industries: Develops advanced defense technologies integrating AI and autonomous systems to enhance national security. Its Lattice platform powers a family of systems that provide real-time, 3D command and control by processing thousands of data streams, enabling capabilities such as counter-unmanned aircraft systems (CUAS) and force protection across land, sea, and air.
Avathon: Provides an industrial AI platform designed to optimize operations in heavy industries, enhancing efficiency and resilience. Its solutions aim to extend the life of critical infrastructure and advance the journey toward autonomy.
BCD iLabs: Develops AI-driven R&D platforms tailored for the food and beverage industry, aiming to accelerate product development cycles and reduce the number of experiments required. Its Innov8 OSplatform enhances product velocity by streamlining formulation and processing.
BrainBox AI: Develops AI-driven HVAC optimization solutions for building management, aiming to reduce energy consumption and greenhouse gas emissions. Its technology leverages deep learning algorithms to predict building energy needs and automate HVAC systems.
causaLens: Specializes in Causal AI, offering a platform that goes beyond traditional machine learning by understanding cause-and-effect relationships. This approach enhances decision-making processes across various industries.
Chemical.AI: Focuses on AI solutions for the chemical industry, providing tools that assist in chemical synthesis planning, reaction prediction, and process optimization to accelerate research and development.
Composabl: Offers a no-code platform for creating industrial-strength autonomous AI agents capable of making high-impact decisions in real-world scenarios. Its technology integrates perception, reasoning, and intuition, enabling AI agents to perform complex tasks alongside human operators.
Edge Impulse: Offers a development platform for machine learning on edge devices, enabling industries to create intelligent solutions that operate directly on hardware with limited resources, enhancing real-time decision-making.
Figure: Specializes in AI-driven solutions for industrial applications, focusing on predictive maintenance, quality control, and process optimization to improve operational efficiency and reduce downtime.
Kelvin : Provides an industrial AI platform that integrates human expertise with machine intelligence to optimize complex industrial operations, aiming to improve efficiency, safety, and sustainability.
ketteQ: Delivers supply chain planning and execution solutions powered by AI, focusing on providing real-time visibility, scenario planning, and optimization to enhance supply chain resilience and efficiency.
Leela AI: Develops AI solutions tailored for industrial applications, focusing on predictive maintenance, quality control, and process optimization to improve operational efficiency and reduce downtime.
Luffy AI: Specializes in AI-driven robotics solutions, providing adaptive control systems that enable robots to learn and adapt to complex tasks in industrial settings, enhancing automation capabilities.
minds.ai: Offers AI solutions for complex system optimization, including applications in automotive design and industrial processes, utilizing deep reinforcement learning to improve performance and efficiency.
parabole.ai: Provides AI-driven solutions for unstructured data processing, enabling industries to extract actionable insights from large volumes of text and documents, enhancing decision-making and operational efficiency.
Physical Intelligence: Aims to bring general-purpose AI into the physical world by developing adaptable AI software for robots. Its mission is to create foundation models capable of controlling any robot to perform any task, enhancing the versatility and applicability of robotics across various industries.
Retrocausal: Develops AI-powered solutions for manufacturing, focusing on real-time error detection and process optimization to improve quality control and reduce operational costs.
SKAIVISION: Offers AI-based computer vision solutions for industrial applications, enabling real-time monitoring, defect detection, and process automation to enhance productivity and quality.
Salus Technical: Provides software solutions that combine AI with engineering expertise to improve process safety and risk management in industrial operations, aiming to prevent accidents and ensure compliance.
Sight Machine: Delivers a Manufacturing Data Platform that utilizes AI to convert unstructured plant data into a standardized data foundation. Its platform continuously analyzes all assets, data sources, and processes to improve production efficiency and enable data-driven transformation in manufacturing.
Traction Ag: Specializes in AI-driven solutions for the agricultural sector, offering tools for crop monitoring, yield prediction, and farm management to enhance productivity and sustainability.
TwinThread: Delivers an AI-powered platform for industrial operations, focusing on predictive operations and performance optimization to improve efficiency, reduce downtime, and enhance decision-making.
Vention: Provides a cloud-based platform that leverages AI to enable the design and deployment of automated equipment, simplifying the automation process for manufacturing industries.

Significant Investment: AI startups have attracted substantial investments, highlighting their importance in the tech landscape. The companies on the Forbes AI 50 list have raised a total of $34.7 billion in funding. This influx of capital enables startups to innovate and scale their operations quickly.

Large Investments in AI Research Firms: Significant funding has gone to AI research firms. For example, OpenAI has received $11.3 billion in funding, and Anthropic has raised $7.7 billion.

Rapid Market Growth: The AI sector is witnessing rapid expansion, evidenced by the increasing number of submissions for awards like the Forbes AI 50 list, which nearly doubled in a single year. This growth underscores the dynamism and competitiveness of the AI market. For the Forbes AI 50 list, approximately 1,900 submissions were received, with a rigorous process that combined quantitative analysis with qualitative evaluations by judges.

AI startups are pivotal in driving the Industrial AI Revolution, acting as agile and innovative forces that bring cutting-edge solutions to the market. Their focused approach, coupled with the significant investments they attract, is fostering the rapid growth of a new tech economy. Their efforts are not only disrupting established markets but also pushing the boundaries of what is possible in industrial automation and setting the stage for a future where AI is seamlessly integrated into various industrial processes.

Industrial-grade AI Battlefronts: Where the Rubber Meets the Road

Within the larger “AI Wars”, specific industrial needs are creating their own battlefronts, and alliances.

Industrial-grade Data Scientists: The demand for AI experts who also understand the nuances of manufacturing and industrial processes is growing. This is a recognized need, as evidenced by the focus on building in-house expertise with Industrial AI Centers of Excellence (CoE). ARC found evidence in 2024 that Leaders are “widening the digital divide” by building in-house expertise with an Industrial AI CoE, to attract, train and retain “industrial grade” data scientists.
Domain Expertise and Neutrality: Industrial organizations prefer to partner with companies that can bring domain expertise to AI. This was demonstrated by Microsoft’s partnerships with Bayer, Cerence, Rockwell Automation, Saifr, Siemens, and Sight Machine. These companies provide industry-specific models in the Azure AI model catalog.
Industrial-grade Data Fabrics are another battlefront. ARC recommends that mainstream and laggards close the gap with industrial AI leaders by prioritizing investments in the Industrial Data Fabric foundations needed for all AI use cases.
Digital Twins are a low priority for many industrial organizations, despite their potential value. ARC believes that creating the underlying Industrial Data Fabric needed for industrial AI, and the benefits Generative AI will bring to interacting with complex systems will lay necessary foundations that have held back meaningful progress on industrial metaverses.
Partnerships are Key: Industrial organizations are partnering with automation and software vendors, as well as cloud hyperscalers as the new ecosystem for the Industrial AI (R)Evolution takes shape with intense competition for the aforementioned data scientists and industrial domain experts needed to advance industrial AI use cases at scale. The flurry of partnership announcements will likely intensify in 2025.
Chief AI Officers (CAIOs) are becoming more prominent, driving the vision and strategy for AI implementation within organizations. Listen to my conversation with Philippe Rambach, CAIO for Schneider Electric, explaining his role: “SPARC: The Emergence of the Chief AI Officer”.

AI Lobbyist Campaigns: Shaping the Market Through Influence and Policy

The battle for influence and policy shaping is an ongoing part of the AI landscape, with companies actively seeking to shape the development and deployment of AI. This includes efforts to drive adoption by emphasizing data security and privacy, while also attempting to fend off potentially restrictive government legislation.

Microsoft is actively addressing ethical AI adoption and data security through several initiatives:

Updates to Azure AI assist with governance, risk, and compliance workflows, underscoring the need to manage ethical AI adoption.
The Copilot Control System provides data protection, management controls, and reporting to help IT departments adopt and measure the business value of AI and agents.
Microsoft Purview offers tools for data loss prevention and insider risk management, highlighting the importance of data security and privacy in the age of AI. These tools help organizations prevent data oversharing, detect risky AI usage, and ensure that sensitive data is not processed inappropriately.

These actions reflect a broader industry trend toward establishing formal procedures for reviewing and approving AI investments, as noted in ARC Advisory Group Research.

AWS, Google, and OpenAI are also engaged in shaping the AI market through various efforts:

AWS emphasizes the security and privacy of its AI services and offers tools and services that help customers maintain control over their data.
Google is committed to developing AI responsibly, with a focus on safety, security, and privacy. Google’s commitment to developing AI responsibly is highlighted in its AI Principles, which also address the societal impacts of AI. Google’s Cloud AI services are designed with enterprise-grade governance, security, and data privacy built-in.
OpenAI has been promoting AI safety and responsible AI development, updating its OpenAI o1 system card to highlight safety improvements and red teaming insights.

These tech companies also engage with governments and regulatory bodies to influence policy decisions related to AI. This includes participating in public consultations, offering recommendations, and advocating for policies that encourage AI innovation while also addressing ethical concerns.

ARC Advisory Group analysts emphasize the need for a Governance Council for ethical and inclusive AI, with global, multi-disciplinary teams that include IT, OT, ET, Workforce, and ESG stakeholder representation. This is a recommendation that all companies should adopt.

Government Legislation in the AI Space

Governments worldwide are actively legislating to ensure that they get a share of the AI action, and that AI development and deployment align with their national priorities. This reflects a growing recognition of the strategic importance of AI and the need to regulate its use.

Regulatory Frameworks: Governments are implementing stringent regulations to ensure the ethical and responsible use of AI. These regulations address issues such as data privacy, algorithmic bias, and the potential impact of AI on employment and society.

Focus on AI Safety and Security: There is a growing emphasis on AI safety and security, with governments focusing on ensuring AI systems are robust and resilient to cyber threats.

The National Institute of Standards and Technology (NIST) has released the NIST AI Risk Management Framework, underscoring the importance of managing risks associated with AI technologies.
Governments are also targeting testing and validation of “Frontier” AI models whose massive cost and scale adoption could be disruptive if not ethically trained, accurate, and explainable before market deployment.

Data Sovereignty: Governments and organizations are competing for control over their data, recognizing its strategic value in powering AI systems. This has led to discussions and policies around data localization, ensuring that data generated within a country remains within its borders, and a focus on the use of local models trained on local data.

Investment and Incentives: Governments are also investing in AI research and development and offering incentives to companies that develop AI technologies. Many governments see AI as critical for economic growth and national security.

International Cooperation: There is ongoing dialogue and collaboration between countries to harmonize AI regulations and address global challenges. These efforts aim to create a more consistent and predictable regulatory environment for AI development and deployment.

The interplay between industry and government is a dynamic and critical aspect of the AI landscape. While companies like Microsoft, AWS, Google, and OpenAI seek to drive adoption through ethical and secure practices, governments are actively shaping the legal and regulatory environment to balance innovation with societal needs. This continuous dialogue will shape how AI is developed, deployed, and utilized in the years to come.

The AI Wars are Just Getting Started

The AI Wars are still in their infancy, and the events of 2024 have set the stage for further advancements and intense competition in the years to come. Here are some ARC Advisory Group predictions for the near future:

From PoCs to Scale: We expect to see a major shift from proof-of-concept AI projects to scaled deployments as the accuracy of foundation models increases, distillation techniques improve, and smaller, more specialized models become more prevalent.
Edge AI will be Key: The value of Edge AI will continue to increase as smaller, more capable inference hardware becomes available.
Data & AI Tech Stack Productivity: We will see continued investments in more productive data and AI technology stacks with multi-agent collaboration and orchestration capabilities.
Business Outcomes: As the range of industrial AI use cases that can deliver positive business outcomes broadens, we will see continued deployments at both the industrial edge and the enterprise cloud.

The AI Wars analogy is a useful tool for making sense of a complex and fast-moving landscape. As we move into 2025, the battle lines are drawn, and the race to capture the benefits of AI is well underway. It is not just a race for technology supremacy—it is also a race to ensure that AI serves humanity with ethical and sustainable outcomes.

The post The AI Wars: Battlefronts, Breakthroughs, and the New Era of the Industrial AI (R)Evolution appeared first on Logistics Viewpoints.

Non classé

The AI Trade Just Ran Into Its Unit-Economics Problem

Published

15 heures ago

9 septembre 2026

The artificial intelligence debate is changing.

For several years, the central question was whether AI worked. Then it became whether the technology could scale.

Increasingly, the question is becoming more difficult:

Who actually captures the value?

That distinction matters because AI adoption can accelerate, token consumption can soar and the economics of individual participants can still deteriorate.

This is the point at which a technology story becomes an operating-model story.

And for logistics companies deciding where AI creates economic advantage, that may be much more useful than watching the daily valuation of another AI stock.

More Tokens Does Not Automatically Mean More Revenue

The basic AI economic equation has generally seemed straightforward.

Models become better. Better models create more useful applications. Applications increase demand. Demand consumes more tokens. More tokens create more revenue.

The complication is price.

AI inference is becoming cheaper at a remarkable rate. Competition among model providers is increasing, open-weight models are improving, specialized models are proliferating and hardware efficiency continues to advance.

That is excellent for customers.

It is less obviously excellent for every company selling tokens.

Man Group’s bearish analysis of the AI investment cycle describes the problem starkly: if token prices decline faster than inference demand expands, enormous growth in usage does not necessarily produce the revenue required to justify an equally enormous infrastructure buildout.

Goldman Sachs Research is more constructive. Its work points to sharply increasing token consumption – potentially a 24-fold increase by 2030 as agentic AI expands – while declining unit compute costs could improve hyperscaler margins.

Those positions are not actually contradictory.

They describe the variable that matters.

The economic outcome depends on the relationship among volume, price and cost.

That is unit economics.

The De-Rating Is Telling Us Something

The financial market has already become more discriminating about AI exposure.

A recent Goldman basket analysis reportedly showed an all-inclusive AI pair roughly 46 percent below its previous highs, even as underlying demand for compute remains strong. Goldman’s argument is not that AI is ending; rather, the opportunity may be broadening toward businesses with clearer monetization, embedded workflows and defensible economic positions.

That distinction is important.

A broad technology transition does not guarantee that every participant in the technology stack earns extraordinary returns.

The internet changed the world. Many internet companies disappeared.

Containerization transformed global trade. That did not mean every shipping line earned exceptional margins.

Cloud computing became foundational infrastructure. The economics nevertheless concentrated in particular layers of the stack.

AI is likely to behave similarly.

The technology can be revolutionary while the value migrates.

Open Models Are Accelerating the Pressure

The open-model transition makes this more visible.

Vercel’s AI Gateway data recently showed open-weight models reaching approximately 62 percent of token volume on August 22, up from 28.4 percent roughly two months earlier and about 11 percent in April. These are figures from one platform rather than the entire AI industry, but the speed of the shift is notable.

Customers are learning something logistics managers learned long ago.

Not every movement requires premium service.

You do not ship every load by air. You do not give every SKU the same inventory policy. You do not assign the most expensive resource to every task simply because it is technically capable of performing it.

The same logic is beginning to apply to AI.

Enterprises can route difficult tasks to expensive frontier models while sending repetitive, lower-value or highly specialized workloads to cheaper models.

This is good architecture.

It is also price pressure.

As switching becomes easier, the model itself can become one component within a larger decision architecture.

That changes where the economic moat resides.

The Value May Move Up the Stack

Consider a logistics company using AI to manage exceptions.

The model may analyze late shipments, weather, inventory availability, customer commitments and transportation alternatives. It may recommend that an order be reallocated, a shipment expedited or a customer promise changed.

Suppose the model call costs 20 cents today and two cents three years from now.

The model provider has experienced severe unit-price compression.

The logistics operator has not necessarily lost anything.

In fact, the logistics operator may gain.

If the AI avoids a $5,000 expedite, prevents a stockout or allows one planner to manage twice as many exceptions, the economic value exists primarily in the operating outcome – not in the token.

This is where the AI economics discussion becomes particularly relevant to enterprise logistics.

Falling model prices could transfer economic value away from model providers and toward companies capable of embedding inexpensive intelligence into valuable workflows.

The model becomes cheaper.

The decision becomes more valuable.

Architecture Becomes the Moat

This suggests that enterprises should be careful about defining an “AI strategy” around access to a particular model.

Model leadership can change.

Prices can change even faster.

An enterprise architecture built tightly around a single provider may eventually look like a transportation network designed around one carrier regardless of lane, service requirement or price.

The more durable architecture is likely to separate the business problem from the intelligence resource used to solve it.

Understand the decision.

Assemble the required context.

Define the constraints.

Determine the acceptable action.

Then route the problem to the model – or combination of models – capable of solving it at the appropriate cost.

That is not simply AI adoption.

It is AI orchestration.

And it looks very similar to other logistics optimization problems.

Intelligence Is Becoming a Variable Cost

The larger economic change may be that machine intelligence is becoming a purchasable input whose price declines rapidly.

That is extraordinary.

For most of industrial history, intelligence has been expensive. Analytical capacity was constrained by the number of skilled people available to perform the work.

AI begins to relax that constraint.

If the cost of a useful unit of machine reasoning continues falling, companies can economically apply intelligence to thousands of decisions that previously could not justify human analysis.

Shipment prioritization.

Carrier selection.

Inventory rebalancing.

Appointment scheduling.

Exception resolution.

Warehouse labor planning.

Supplier-risk analysis.

Customer-response generation.

Every one of these can potentially consume enormous quantities of tokens while producing economic value far larger than the cost of those tokens.

This is why collapsing AI prices are not necessarily bearish for AI adoption.

They may be extraordinarily bullish for the users of AI.

The Second Half of the AI Trade

The first phase of the AI boom rewarded scarcity.

There were too few advanced GPUs. Too little compute. Too few frontier models. Too little capacity.

Scarcity created pricing power.

The next phase may reward orchestration.

Models will proliferate. Intelligence will become cheaper. Enterprises will learn to switch among providers. Open-weight systems will handle an increasing share of routine workloads. Infrastructure costs will continue to matter, but customers will become much more disciplined about what they are willing to pay for a unit of intelligence.

That does not mean the AI boom is over.

It means the economics are maturing.

The winners in logistics will probably not be the companies that consume the most AI.

They will be the companies that turn increasingly inexpensive intelligence into better operating decisions.

That is a very different metric.

Tokens are an input.

The decision is the product.

And the economic value ultimately belongs to whoever can make that decision improve the system.

The post The AI Trade Just Ran Into Its Unit-Economics Problem appeared first on Logistics Viewpoints.

Non classé

Requirements Before Technology: Define the Problem Before Buying the Solution

Published

16 heures ago

9 septembre 2026

The fastest way to buy the wrong logistics technology is to begin with the technology. Yet that is still how too many programs start: a new TMS, WMS, control tower, yard platform, AI initiative, or automation project. The solution enters the room before the operating requirement has been defined.

The problem is not that any of these technologies are bad ideas. The problem is that the solution has entered the conversation before the requirement has been defined. Systems engineering reverses that order. The expanding scope described in What Is a WMS in 2026? is a useful example of why requirements should be defined before a buyer lets a rapidly broadening product category define the problem.

Start With the Operating Problem

A requirement is not a feature request. “AI-enabled routing” is not a business requirement. “Reduce the time required to identify and recover priority shipments at risk of missing their delivery commitment” is closer. “Real-time visibility” is not a requirement by itself. “Identify at-risk customer orders early enough to take corrective action before the promised delivery window” is.

The difference matters because requirements describe desired system behavior and outcomes. Features describe how a vendor has chosen to implement capabilities. When organizations begin with features, the evaluation tends to become a comparison of product checklists. When they begin with requirements, they can evaluate whether a technology, process change, organizational change, or combination of solutions actually solves the operating problem. That produces a very different buying process.

Logistics requirements are rarely one-dimensional. At the highest level, the business may require improved service, lower working capital, greater resilience, faster response, or lower operating cost. Those outcomes then need to be translated into more specific operating requirements.

If the objective is faster response to disruption, what does faster mean? Minutes, hours, or days? Which disruptions matter? What information must be available? Which decisions must be accelerated? Who is allowed to make them? What constraints cannot be violated?

The answers drive system design. A useful requirements hierarchy might include business requirements, operational requirements, information requirements, decision requirements, technology requirements, security and compliance requirements, and human factors requirements. That last category is easy to neglect. A system may be technically capable of producing a recommendation every five minutes, but if a planner can realistically evaluate only ten exceptions per hour, the operating design still has a bottleneck.

Put the Constraints on the Table Early

Good requirements engineering does more than define what a system should do. It defines the boundaries within which the system must operate. Logistics networks are full of constraints: labor availability, warehouse throughput, dock capacity, trailer availability, carrier schedules, driver hours, parcel cutoffs, regulatory requirements, customer commitments, data limitations, capital budgets, integration dependencies, physical space, maintenance windows, and organizational policy. If those constraints are left implicit, they eventually reappear as implementation surprises.

This is particularly important in automation and AI projects. Optimization models are only useful when they reflect the constraints that actually govern the operation. AI recommendations are only actionable when they fit within decision rights, data quality, and execution capability. The best technology in the world cannot compensate for a requirement that was never articulated. Requirements discussions also benefit from distinguishing between what is necessary and what is desirable.

Every stakeholder has preferences. Transportation may want a particular carrier workflow. Finance may want additional controls. IT may prefer a specific architecture. Operations may want a familiar user interface. Executives may want a capability they have seen elsewhere.

Some of these preferences are important. Others are habits. A disciplined process identifies which requirements are mandatory, which are high-value, which are negotiable, and which are simply convenient. That distinction gives the organization room to make intelligent tradeoffs rather than creating an impossible specification in which everything is equally important.

It also improves vendor conversations. Technology and automation providers can respond to the logistics problem the customer is actually trying to solve rather than to an undifferentiated list of requests. One of the most useful ideas from systems engineering is that a requirement should eventually be verifiable. “Improve visibility” is difficult to test.

“Provide the status and predicted arrival time of 95 percent of priority inbound shipments with data no more than 30 minutes old” can be tested. This discipline creates a bridge between design and implementation. The requirements used to justify the investment become the basis for validation after the system is deployed.

That sounds elementary, but many logistics projects lose this connection. Business cases are approved around outcomes, implementations are managed around milestones, and success is eventually declared because the system went live. Go-live is not a business outcome. A system should be judged against what it was supposed to accomplish.

Buy the Solution Only After the Problem Is Defined

A requirements-first approach does not slow innovation. It makes innovation more precise. Once the operating requirements are clear, technology choices become easier to evaluate. The organization can determine which capabilities are essential, which integrations matter, where automation is appropriate, where human judgment remains important, and what level of performance is actually required.

Sometimes the answer will be a new platform. Sometimes it will be process redesign, better master data, a change in decision rights, or a relatively modest extension of an existing system. That is not a less ambitious transformation. It is a more engineered one.

Logistics leaders are under enormous pressure to move quickly, especially around AI and automation. Speed matters. But speed in selecting a solution is not the same as speed in solving the problem.

Speed matters, especially in AI and automation. But speed in selecting a solution is not the same as speed in solving the problem. In 1.4, we take the next step and make the stakeholder conflicts, constraints, and tradeoffs explicit before they get buried in the design.

Related Logistics Viewpoints research

Systems Engineering in Logistics
The New Architecture of Logistics
2026 Supply Chain Decision Intelligence Market Map
Supply Chain Technology Buyers Have a Market Structure Problem
Previous in this series: Stop Managing Logistics as a Collection of Functions

Request the Systems Engineering in Logistics Client Edition

If your organization is evaluating a logistics transformation, technology strategy, automation program, or operating-model redesign, I would be glad to provide the complete client edition and discuss how the framework applies to your priorities, constraints, and operating environment.

Request the client edition

The post Requirements Before Technology: Define the Problem Before Buying the Solution appeared first on Logistics Viewpoints.

Non classé

NVIDIA Is Buying the Distribution Layer of AI

Published

1 jour ago

8 septembre 2026

NVIDIA’s agreement to acquire Hugging Face for approximately $12.9 billion looks, at first, like another large transaction in an AI market already full of large numbers. Look more closely, however, and this is considerably more interesting than a semiconductor company buying a software company.

NVIDIA already dominates one of the most important layers of artificial intelligence: accelerated computing. Hugging Face occupies a different position. It has become one of the principal places where developers discover models, evaluate them, modify them, and decide how and where those models should run.

NVIDIA is therefore not simply acquiring another AI asset. It is moving toward the interchange where models, applications, developers, and computing infrastructure meet. That matters because the next phase of AI competition will increasingly be about orchestration rather than individual components.

For logistics executives, that should sound familiar.

Hugging Face Has Become AI Infrastructure

Hugging Face began in 2016 and has evolved into much more than a repository for AI models. The platform now serves millions of developers and hosts millions of models, along with datasets and applications used by companies to discover, evaluate, customize, and deploy artificial intelligence.

The scale matters, but its position within the architecture matters more.

Modern AI is increasingly becoming a component ecosystem. An enterprise does not necessarily select one enormous model and build everything around it. It might use one model for computer vision, another for document processing, another for coding, and a more capable frontier model for difficult reasoning. Some models might run internally, others through cloud APIs, and still others at the edge.

Hugging Face sits in the middle of that increasingly complicated environment. It helps developers find the components and increasingly helps them determine how those components can be deployed.

In logistics terms, Hugging Face looks less like a manufacturer and more like an interchange. It does not need to manufacture every product moving through the network to influence how the network operates. Its value comes from connecting a large number of models, developers, applications, and infrastructure choices.

NVIDIA has agreed to buy that interchange.

NVIDIA’s Problem Is Bigger Than GPUs

NVIDIA’s existing position in artificial intelligence is extraordinary, but it also contains a strategic vulnerability. Some of its largest customers have powerful incentives to reduce their dependence on NVIDIA hardware. Microsoft, Google, Amazon, Meta, OpenAI, and others have developed or are developing specialized accelerators of their own.

That does not mean NVIDIA’s GPU franchise disappears. Its installed ecosystem, software architecture, developer expertise, and performance advantages remain formidable. But it does mean NVIDIA cannot assume the future AI architecture will consist of NVIDIA hardware underneath every important workload.

The rational response is to expand the battlefield.

If AI infrastructure becomes increasingly heterogeneous, then the layer that helps determine what models are selected and where workloads are executed becomes more valuable. NVIDIA does not necessarily have to own every model or manufacture every accelerator if it can remain deeply embedded in the architecture through which the larger ecosystem operates.

Hugging Face provides a route into that architecture.

A developer might choose a model created by Meta, Mistral, Google, DeepSeek, or an independent research group. That model might eventually run on NVIDIA hardware, an AMD accelerator, a hyperscaler’s custom chip, an enterprise server, or an edge device.

Owning Hugging Face puts NVIDIA much closer to the point where those decisions originate.

Why Hugging Face Needs to Remain Open

One of the most revealing parts of the deal is NVIDIA’s commitment to keep Hugging Face open and compute-agnostic. NVIDIA has said its hardware will not be required to build or deploy through Hugging Face and that the platform will continue supporting multiple clouds, frameworks, models, inference providers, and silicon architectures.

That may sound counterintuitive. Why spend almost $13 billion on a platform and continue allowing competing hardware through it?

Because the neutrality of the platform is part of what makes it valuable.

An interchange becomes strategically important because many participants are willing to use it. Ports become powerful because multiple carriers call there. Freight marketplaces become more valuable as more shippers and carriers participate. Digital platforms acquire influence because participants on multiple sides of a market continue to meet there.

If NVIDIA turned Hugging Face into a closed distribution channel for NVIDIA hardware, it could weaken the network effect it is buying.

The better strategy is subtler. Keep the interchange open, encourage as much AI development and deployment as possible to flow through it, and make NVIDIA infrastructure exceptionally attractive when users decide where those workloads should run.

That is not exclusivity. It is influence.

AI Is Becoming a Routing Problem

The acquisition also arrives as enterprise AI moves away from the assumption that the largest available model should handle every task.

That makes little economic sense once organizations begin operating AI at scale.

Logistics has dealt with this type of problem for decades. A company does not ship every product by air because air freight is fast. It selects the mode appropriate to the service requirement, product value, distance, urgency, and cost. The fastest resource is not necessarily the economically correct resource.

AI workloads are beginning to require the same discipline.

A difficult planning problem involving incomplete information may justify a high-cost frontier reasoning model. Invoice classification probably does not. A computer-vision application may require something entirely different. A repetitive enterprise workflow may run perfectly well on a smaller open-weight model hosted internally at a fraction of the cost.

Once enterprises begin making these choices systematically, model selection becomes a routing problem. Capability, cost, latency, reliability, privacy, sovereignty, and infrastructure availability all become constraints.

Hugging Face occupies an important position in that emerging routing architecture because it provides access to a broad universe of models rather than forcing developers into a single supplier ecosystem.

NVIDIA is now buying a position much closer to that routing decision.

From Components to Systems

The acquisition fits a broader evolution in NVIDIA’s strategy. The company has been steadily expanding outward from the GPU into networking, software libraries, complete computing systems, inference infrastructure, robotics, digital twins, and AI factories.

Hugging Face adds another layer: developers, models, datasets, applications, and distribution.

Viewed as a system, the logic becomes clearer. At the bottom of the architecture, NVIDIA supplies much of the physical computing infrastructure. Higher in the stack, its software and development tools help applications use that infrastructure. Hugging Face gives the company a strategic position closer to where developers choose which models to use and how those models should be deployed.

That means NVIDIA does not need every workload in the ecosystem to run on NVIDIA hardware for the acquisition to succeed. The larger objective may be to grow the entire AI ecosystem while positioning NVIDIA infrastructure as one of the easiest and most attractive destinations for the resulting workload.

That is a platform strategy, and it is considerably more durable than a strategy based solely on hardware scarcity.

The Logistics Lesson

There is a broader lesson here for logistics because the same structural shift is occurring throughout industrial technology.

Companies naturally focus on assets: factories, warehouses, transportation capacity, automation equipment, software applications, semiconductors, and increasingly AI models. But as systems become more interconnected, an increasing share of competitive power moves into the interfaces between those assets.

The valuable position is often the place where choices are made.

Which carrier receives the load? Which warehouse fills the order? Which inventory pool serves the customer? Which model handles the request? Which computing resource executes the workload?

The organization that controls or intelligently orchestrates those decisions can acquire influence far beyond the value of the underlying asset.

This is why orchestration is becoming such an important theme across logistics. Companies have spent decades improving individual nodes. They now have better warehouses, better transportation systems, better planning applications, better automation, and better visibility. The next increment of performance increasingly comes from coordinating those resources as a system.

Artificial intelligence is following the same path.

The first stage of the AI boom was dominated by the question of who could build the most powerful individual components. NVIDIA won an extraordinary portion of that contest because its GPUs became the essential machinery of AI.

The next contest will be about how those components are selected, routed, and orchestrated.

That is what makes Hugging Face strategically important. NVIDIA already controls one of the most valuable resources in artificial intelligence. It is now buying a position much closer to the place where millions of developers decide what intelligence to use and how to deploy it.

The chips still matter enormously, but the center of gravity is moving from individual components toward the architecture connecting them. As logistics has demonstrated repeatedly, the company that controls the interchange can become just as important as the companies producing what moves through it.

The post NVIDIA Is Buying the Distribution Layer of AI appeared first on Logistics Viewpoints.