Load forecasts from data centers risks falling into irrational exuberance territory

Stay Ahead, Stay ONMINE

Load forecasts from data centers risks falling into irrational exuberance territory

Todd Snitchler is president and CEO of the Electric Power Supply Association. Alan Greenspan is quoted as saying in 1996 that the stock market experienced irrational exuberance around the dot com bubble when describing the disconnect between asset valuation and a stock’s value rising to prices that exceed their fundamentals. In many ways, the race […]

Todd Snitchler is president and CEO of the Electric Power Supply Association.

Alan Greenspan is quoted as saying in 1996 that the stock market experienced irrational exuberance around the dot com bubble when describing the disconnect between asset valuation and a stock’s value rising to prices that exceed their fundamentals. In many ways, the race to power our artificial intelligence-driven future is showing the same thing. There is no doubt that the growth of AI will be driven or limited by how quickly we can ensure sufficient power to enable the expansion and operation of data centers, but the magnitude of the energy needed is still much more in question than many seem to think.

The recent announcements by PJM Interconnection of their load forecast indicate a significant increase in demand. For summer demand, they predict an increase of 42 GW between now and 2029 and, for winter demand, an increase of 43 GW by 2039. The incremental growth expected by 2034 is notable with a 27 GW and 30 GW increase in summer and winter peak demand.

The North American Electric Reliability Corp. also released its 2024 Long-Term Reliability Assessment which projects a 15% summer peak demand increase, and an 18% winter peak demand increase over the next decade. While these timelines and figures have some variance, they all suggest a consistent, meaningful level of load growth in all seasons of the year in large parts of the country.

There is also a long history of overestimating load growth which can drive overbuilding and the addition of resources that may exceed the increased demand. A similar panic about load growth overwhelming the grid happened back in 1999, when worries about early data centers led to panicked media stories and rosy investor forecasts for utilities. It didn’t happen. Microprocessors, computers and data storage saw huge gains in efficiency that kept pace with their gains in raw power and the predicted load growth never materialized.

Load projections compared to actual load growth since 2006 reveal that estimates are often wildly optimistic compared to the actual demand the system achieved over time.

What is the lesson we need to heed? This does not mean that load growth is not coming or that we do not have an electricity supply issue that needs prompt attention. What it does mean, though, is that we need to be judicious in our thinking, methodical with our planning, and focused on optimizing the system performance.

A careful analysis of the actual availability of high-end processing chips over the next several years shows that they are already “sold out,” which means there is a limit on the ability of the AI industry and data centers to grow exponentially and overconsume energy, stressing the electrical grid to the point of failure.

This fact will help inform the speed and volume of likely data center development regardless of the region of the country where development is projected. Advancements in AI may also play a role in making the grid itself more efficient. Additional chip manufacturing capacity is subject to the speed to market and time it takes to construct and operate any new fabricating capacity.

The tech industry has also acknowledged that demand is far from certain. In a recent filing in Ohio, Amazon Web Services expert Michael Fradette told regulators that asking companies to predict how much power data centers will need over the next decade with a high degree of accuracy is “unreasonable,” because consumption will depend on factors including customer demand and emerging technologies.

Too many families are already struggling with high energy prices. A historic buildout of new generation resources and a binge of utility spending to meet this forecasted demand would be prohibitively expensive for consumers.

Major investments will be needed, but it is incumbent on policymakers to be clear-eyed about the costs and consequences. State and national regulators should be careful to avoid overbuilding, eager to embrace cost-effective and market-based solutions, and committed to ensuring customers are not the ones carrying the burden of investments that don’t turn out to be needed.

Competitive power markets are a key tool at our disposal to make sure those risks and costs do not negatively impact consumers, by shifting that risk from utility customers to shareholders. A more streamlined permitting system can also drastically reduce the costs and time to build new infrastructure without adding costs for consumers.

America’s past experience with ambitious load forecasts would caution against overexuberance and instead encourage a rational, deliberate build out of resources needed to power the evolving American economy as we move deeper into the 21st century.

Stay Ahead

Explore More Insights

Stay ahead with more perspectives on cutting-edge power, infrastructure, energy, bitcoin and AI solutions. Explore these articles to uncover strategies and insights shaping the future of industries.

MAD Chairs: A new tool to evaluate AI

This is a linkpost for https://arxiv.org/abs/2503.20986v2

IBM Cloud speeds AI workloads with Intel Gaudi 3 accelerators

For businesses that need more control over their AI development, IBM says they can deploy IBM watsonx.ai software with the Intel Gaudi 3-based virtual server on IBM Cloud VPC in Q2 2025. IBM watsonx.ai includes an end-to-end AI development studio, AI developer toolkit and full AI lifecycle management for developing AI services

Nvidia’s Blackwell raises the bar with new MLPerf Inference V5.0 results

Fifteen partners, including Cisco, Fujitsu, Hewlett Packard Enterprise, Dell Technologies, Oracle, and Google Cloud, were involved in the latest round of MLPerf testing, which, he noted, was the largest number of Nvidia partners submitting to the benchmark in any given round. When asked about his overall impression on the latest

Palo Alto VP Jordi Botifoll: ‘You can’t play with cybersecurity’

Palo Alto has boosted this effort in recent years with the integration of precision artificial intelligence that includes machine learning and deep learning techniques, in addition to generative AI tools. “Our strategy is to ensure that the time to detect an attack and the time to resolve it (if it

BP, Repsol Start Production at Cypre Gas Project in Trinidad

BP PLC said Monday the Cypre field in Trinidad and Tobago is now producing and is expected to deliver about 250 million standard cubic feet a day (MMscfd) of gas at peak. The project, located 78 kilometers (48.47 miles) off the southeast coast of Trinidad in a water depth of about 80 meters (262.47 feet) according to BP, is part of the East Mayaro Block. Cypre is wholly owned by BP Trinidad and Tobago LLC (BPTT), a joint venture owned 70 percent by Britain’s BP and 30 percent by Spain’s Repsol SA. Cypre is one of 10 projects that BP aims to start up between 2025 and 2027. “Production from Cypre will make a significant contribution towards the 250,000 barrels of oil equivalent per day combined peak net production expected from these 10 projects”, it said in an online statement. BPTT’s third subsea development, Cypre will have 7 wells tied back to BPTT’s existing Jupiter platform. Phase 1, consisting of 4 wells, was completed at the end of 2024. “The second phase is expected to commence in the second half of this year”, BP said. BPTT president David Campbell said, “Cypre is another key milestone in bpTT’s strategy to maximize production from our shallow water acreage using existing infrastructure”. “The project not only reinforces our commitment to maintaining production but also plays a crucial role in satisfying our existing gas supply commitments”, Campbell added. “Cypre represents a significant investment in the country’s energy sector”. William Lin, BP executive vice president for gas and low-carbon energy, said, “The second of 10 major projects across our global portfolio that we expect to start up by 2027, Cypre is also the first of a series of projects we will be bringing online in Trinidad to deliver gas to the nation and add value for

USA Crude Oil Inventories Rise More Than 6MM Barrels Week on Week

U.S. commercial crude oil inventories, excluding those in the Strategic Petroleum Reserve (SPR), increased by 6.2 million barrels from the week ending March 21 to the week ending March 28, the U.S. Energy Information Administration (EIA) highlighted in its latest weekly petroleum status report. That report was released on April 2 and included data for the week ending March 28. The EIA report showed that crude oil stocks, not including the SPR, stood at 439.8 million barrels on March 28, 433.6 million barrels on March 21, and 451.4 million barrels on March 29, 2024. Crude oil in the SPR stood at 396.4 million barrels on March 28, 396.1 million barrels on March 21, and 363.6 million barrels on March 29, 2024, the report outlined. Total petroleum stocks – including crude oil, total motor gasoline, fuel ethanol, kerosene type jet fuel, distillate fuel oil, residual fuel oil, propane/propylene, and other oils – stood at 1.605 billion barrels on March 28, the report showed. Total petroleum stocks were up 5.6 million barrels week on week and up 27.2 million barrels year on year, the report revealed. “At 439.8 million barrels, U.S. crude oil inventories are about four percent below the five year average for this time of year,” the EIA said in its report. “Total motor gasoline inventories decreased by 1.6 million barrels from last week and are two percent above the five year average for this time of year. Finished gasoline inventories increased and blending components inventories decreased last week,” it added. “Distillate fuel inventories increased by 0.3 million barrels last week and are about six percent below the five year average for this time of year. Propane/propylene inventories increased by 1.0 million barrels from last week and are eight percent below the five year average for this time of year,”

Kumul Wraps Up Kimu, Barikewa Seismic Program

Papua New Guinea’s national petroleum and energy company Kumul Petroleum Holdings Limited has finished its seismic program on its operated Kimu and Barikewa licenses. “On Wednesday, March 26, we shot the last hole on a seismic line in the Kimu license, and camp demobilization has begun. This is the culmination of a seismic program we have been undertaking in our two petroleum retention licenses”, Wapu Sonk, Kumul Petroleum’s managing director, said. “The purpose of this seismic work is to gather additional geological information on the petroleum resources already discovered in these two licenses, to increase the volumes of oil and gas reserves. The second objective is to identify and firm up siting of some delineation wells and exploration targets”. Sonk said the seismic campaign had been carried out by contractor OilMin Holdings and had taken approximately six months. He noted that seismic line cutting is a labor-intensive activity and the six seismic lines required the employment of 477 local people in remote areas of the Gulf and Western provinces. The company anticipates 6 months for analyzing and incorporating seismic data into its geological models. “Kumul Petroleum and OilMin have worked closely with locally impacted communities and although seismic work only requires company presence in an area for a relatively short time, assistance has been provided to local primary schools at Kaiam and Kumusi. Without community support this fieldwork would not have been possible”, Sonk said. Sonk said Kumul Petroleum is investing heavily to assess and develop its oil and gas licenses, aiming to create economically viable development plans. He highlighted Kumul Petroleum’s direct involvement in exploring and potentially commercializing previously deemed unviable gas resources in Papua New Guinea. To contact the author, email [email protected] What do you think? We’d love to hear from you, join the conversation on the Rigzone

Buru Brings on CEFA to Co-Develop Rafael Gas Project in Australia

Buru Energy Ltd. has executed a Strategic Development Agreement (SDA) with Clean Energy Fuels Australia Pty. Ltd. (CEFA) to co-develop the Rafael gas project. The project is located in the Canning Basin, some 150 kilometers (93.2 miles) east of Broome and approximately 85 kilometers (52.8 miles) south of Derby in the Shire of Derby-West Kimberley, Western Australia. Buru said in a media release the project is targeting the replacement of long-haul trucked or imported fuel used for power generation and mining in the northwest of Western Australia with a local source of trucked liquefied natural gas (LNG) and liquids, supporting the development of new market opportunities in the region. “The agreement with CEFA is a watershed moment for Buru and the Rafael gas project. It marks a clear demonstration of the company’s gas strategy and transition from explorer to developer and long-term producer”, Buru CEO Thomas Nador said. “Rafael is the only confirmed source of conventional gas and liquids in onshore Western Australia north of the North West Shelf project. It is a unique opportunity to provide energy to a growing market that is not connected to a gas pipeline and currently faces challenges with high energy costs and security of supply”. “Our proven virtual pipeline or ‘trucked LNG’ model lowers long-term regional energy costs and emissions and provides a viable alternative to diesel for new and existing energy users”, Basil Lenzo, CEFA/Octa Group CEO and Director, said. Buru and CEFA will establish gas pricing arrangements for competitive LNG sales, ensuring economic returns and value sharing, Buru said. The two companies will also work together on condensate production and sales. Buru added that CEFA will finance and operate a small-scale LNG plant with a capacity of up to 300 tonnes per day on the Rafael 1 well pad, with a

OEUK calls for UK hydrogen scale-up

Offshore Energies UK (OEUK) has called for policymakers to accelerate deployment of the UK hydrogen sector. In a new report, the body warned that the country’s hydrogen industry has been slower to develop than expected and outlined key recommendations to help boost its growth. Among its recommendations are prioritising the deployment of hydrogen into industries where it can make the biggest impact in decarbonising sectors, such as high-temperature manufacturing processes. In addition, OEUK said that the government should commit to realistic, deliverable deployment targets to attract investment and build confidence in the sector. The report also called for ensuring funding mechanisms for hydrogen business models, such as the proposed Gas Shipper Obligation, are fair and proportionate whilst balancing the need to preserve energy security, reduce energy costs and deliver net zero objectives. Finally, OEUK said that policies must recognise the critical importance of domestic gas production in delivering the scale of low carbon hydrogen supplies needed to create this new market. OEUK head of energy policy Enrique Cornejo said: “The UK has the people, projects, and potential to make hydrogen a cornerstone of our energy future – but without urgent action, that opportunity could slip through our fingers. “The UK has over 100 hydrogen projects in the pipeline, equating to more than 15GW of production capacity, but without clarity on key areas and policies, these projects will struggle to progress. “Hydrogen has a role to play as a decarbonisation solution and for those industries, there needs to be a scale up of the sector. This isn’t only about hitting targets – it’s about securing jobs, investment, and a future for many of these sectors.” The UK hydrogen industry recently echoed some of OEUK’s calls, saying that the government needs to deliver “pragmatism” and policy certainty to help the sector reach

The UK’s mineral security: The future of clean energy

Just a few short days ago, the Department of Energy announced that the UK’s greenhouse gas emissions had fallen by 3.5% last year – due in large part to a new record for renewable energy generation, such as wind and solar, which now powers more than 50% of the UK’s electricity supply. Greenpeace UK policy director Doug Parr said it showed “the UK’s efforts to tackle climate change are working,” and called on the government to “rapidly make renewables the backbone of our energy system to lower our bills for good.” The UK’s expansion of renewable energy infrastructure and capacity has certainly proven valuable in recent months, as repeated disruptions have continued to impact the supply of fossil fuels from abroad, and skyrocketing energy prices have underscored the need for greater energy security and independence. However, as the country celebrates this new milestone on the journey to net zero, there are significant obstacles that still lie ahead – such as the lack of long-term financing and the challenges of renewable energy intermittency. However, the most significant complication threatening to undermine the UK’s energy transition is not related to the performance or cost of renewable energy, but rather to the availability of the materials that are needed to build and scale a clean energy system. Critical minerals such as lithium, nickel, cobalt, copper and rare earth elements are essential components of rapidly growing energy technologies – including wind turbines, electricity networks and electric vehicles – and securing a stable and reliable supply of these resources has proven to be an incredibly competitive process. Rare earth boom As of 2025, the rare earth metals market is projected to reach $8.15 billion, growing at a compound annual growth rate of 7.9%. The energy sector’s demand has been a key driver of this surge,

New MLCommons benchmarks to test AI infrastructure performance

The latest release also broadens its scope beyond chatbot benchmarks. A new graph neural network (GNN) test targets datacenter-class hardware and is designed for workloads like fraud detection, recommendation engines, and knowledge graphs. It uses the RGAT model based on a graph dataset containing over 547 million nodes and 5.8 billion edges. Judging performance Analysts suggest that these benchmarks will make it easier to judge the performance of various hardware chips and clusters based on documented models. “As every chipmaker seeks to prove that its hardware is good enough to support AI, we now have a standard benchmark that shows the quality of question support, math, and coding skills associated with hardware,” said Hyoun Park, CEO and Chief Analyst at Amalgam Insights. Chipmakers can now compete not just on traditional speeds and feeds, but in mathematical skill and informational accuracy. This benchmark provides a rare opportunity to add new performance standards on cross-vendor hardware, Park added. “The latency in terms of how quickly tokens are delivered and the time for the user to see the response is the deciding factor,” said Neil Shah, partner and co-founder at Counterpoint Research. “This is where players such as NVIDIA, AMD, and Intel have to get the software right to help developers optimize the models and bring out the best compute performance.” Benchmarking and buying decisions Independent benchmarks like those from MLCommons play a key role in helping buyers evaluate system performance, but relying on them alone may not provide the full picture.

Potential Nvidia chip shortage looms as Chinese customers rush to beat US sales ban

Will it lead to shortages? The US first placed export controls on chips sent to China in October 2022 as a means to slow the country’s technological advances. It blocked the sale of Nvidia’s A100 and H100 chips, leading the company to develop the less powerful A800 and H800 chips for the market; they were also subsequently banned. There was a surge in demand for the H20 following the arrival of Chinese startup DeepSeek’s ultra low-cost, open-source AI model in January. And while the H20 is reported to be 15 times slower than Nvidia’s newest Blackwell chips sold elsewhere in the world, it was designed specifically by Nvidia to comply with the further US export controls introduced in October 2023. It is being used by Chinese companies for training, although it’s billed as an inference chip, explained Matt Kimball, VP and principal analyst for datacenter compute and storage at Moor Insights & Strategy. Should Nvidia choose to focus its efforts on manufacturing more of the chips, Kimball said he doesn’t think it will impact supply in the US and Europe, as Blackwell is the main product sold in those markets and H20 is an N-1 Hopper architecture chip. “If you take this a step further and ask whether this large order slows down the production of chips destined for the US and Europe, I’d say the answer is no, as the Hopper family is built on a different process node than the Blackwell family,” he said. Still, Kimball noted, “supply chain management is difficult, especially for smaller organizations that are put to the back of the line as hyperscalers with multibillion dollar orders are first in line for the newest [chips].”

European cloud group invests to create what it dubs “Trump-proof cloud services”

But analysts have questioned whether the Microsoft move truly addresses those European business concerns. Phil Brunkard, executive counselor at Info-Tech Research Group UK, said, commenting on last month’s announcement of the EU Data Boundary for the Microsoft Cloud, “Microsoft says that customer data will remain stored and processed in the EU and EFTA, but doesn’t guarantee true data sovereignty.” And European companies are now rethinking what data sovereignty means to them. They are moving beyond having it refer to where the data sits to focusing on which vendors control it, and who controls them. Responding to the new Euro cloud plan, another analyst, IDC VP Dave McCarthy, saw the effort as “signaling a growing European push for data control and independence.” “US providers could face tougher competition from EU companies that leverage this tech to offer sovereignty-friendly alternatives. Although €1 million isn’t a game-changer on its own, it’s a clear sign Europe wants to build its own cloud ecosystem—potentially at the expense of US market share,” McCarthy said. “For US providers, this could mean investing in more EU-based data centers or reconfiguring systems to ensure European customers’ data stays within the region. This isn’t just a compliance checkbox. It’s a shift that could hike operational costs and complexity, especially for companies used to running centralized setups.” Adding to the potential bad news for US hyperscalers, McCarthy said that there was little reason to believe that this trend would be limited to Europe. “If Europe pulls this off, other regions might take note and push for similar sovereignty rules. US providers could find themselves adapting to a patchwork of regulations worldwide, forcing a rethink of their global strategies,” McCarthy said. “This isn’t just a European headache, it’s a preview of what could become a broader challenge.”

Talent gap complicates cost-conscious cloud planning

The top strategy so far is what one enterprise calls the “Cloud Team.” You assemble all your people with cloud skills, and your own best software architect, and have the team examine current and proposed cloud applications, looking for a high-level approach that meets business goals. In this process, the team tries to avoid implementation specifics, focusing instead on the notion that a hybrid application has an agile cloud side and a governance-and-sovereignty data center side, and what has to be done is push functionality into the right place. The Cloud Team supporters say that an experienced application architect can deal with the cloud in abstract, without detailed knowledge of cloud tools and costs. For example, the architect can assess the value of using an event-driven versus transactional model without fixating on how either could be done. The idea is to first come up with approaches. Then, developers could work with cloud providers to map each approach to an implementation, and assess the costs, benefits, and risks. Ok, I lied about this being the top strategy—sort of, at least. It’s the only strategy that’s making much sense. The enterprises all start their cloud-reassessment journey on a different tack, but they agree it doesn’t work. The knee-jerk approach to cloud costs is to attack the implementation, not the design. What cloud features did you pick? Could you find ones that cost less? Could you perhaps shed all the special features and just host containers or VMs with no web services at all? Enterprises who try this, meaning almost all of them, report that they save less than 15% on cloud costs, a rate of savings that means roughly a five-year payback on the costs of making the application changes…if they can make them at all. Enterprises used to build all of

Lightmatter launches photonic chips to eliminate GPU idle time in AI data centers

“Silicon photonics can transform HPC, data centers, and networking by providing greater scalability, better energy efficiency, and seamless integration with existing semiconductor manufacturing and packaging technologies,” Jagadeesan added. “Lightmatter’s recent announcement of the Passage L200 co-packaged optics and M1000 reference platform demonstrates an important step toward addressing the interconnect bandwidth and latency between accelerators in AI data centers.” The market timing appears strategic, as enterprises worldwide face increasing computational demands from AI workloads while simultaneously confronting the physical limitations of traditional semiconductor scaling. Silicon photonics offers a potential path forward as conventional approaches reach their limits. Practical applications For enterprise IT leaders, Lightmatter’s technology could impact several key areas of infrastructure planning. AI development teams could see significantly reduced training times for complex models, enabling faster iteration and deployment of AI solutions. Real-time AI applications could benefit from lower latency between processing units, improving responsiveness for time-sensitive operations. Data centers could potentially achieve higher computational density with fewer networking bottlenecks, allowing more efficient use of physical space and resources. Infrastructure costs might be optimized by more efficient utilization of expensive GPU resources, as processors spend less time waiting for data and more time computing. These benefits would be particularly valuable for financial services, healthcare, research institutions, and technology companies working with large-scale AI deployments. Organizations that rely on real-time analysis of large datasets or require rapid training and deployment of complex AI models stand to gain the most from the technology. “Silicon photonics will be a key technology for interconnects across accelerators, racks, and data center fabrics,” Jagadeesan pointed out. “Chiplets and advanced packaging will coexist and dominate intra-package communication. The key aspect is integration, that is companies who have the potential to combine photonics, chiplets, and packaging in a more efficient way will gain competitive advantage.”

Silicon Motion rolls SSD kit to bolster AI workload performance

The kit utilizes the PCIe Dual Ported enterprise-grade SM8366 controller with support for PCIe Gen 5 x4 NVMe 2.0 and OCP 2.5 data center specifications. The 128TB SSD RDK also supports NVMe 2.0 Flexible Data Placement (FDP), a feature that allows advanced data management and improved SSD write efficiency and endurance. “Silicon Motion’s MonTitan SSD RDK offers a comprehensive solution for our customers, enabling them to rapidly develop and deploy enterprise-class SSDs tailored for AI data center and edge server applications.” said Alex Chou, senior vice president of the enterprise storage & display interface solution business at Silicon Motion. Silicon Motion doesn’t make drives, rather it makes reference design kits in different form factors that its customers use to build their own product. Its kits come in E1.S, E3.S, and U.2 form factors. The E1.S and U.2 forms mirror the M.2, which looks like a stick of gum and installs on the motherboard. There are PCI Express enclosures that hold four to six of those drives and plug into one card slot and appear to the system as a single drive.

Microsoft will invest $80B in AI data centers in fiscal 2025

And Microsoft isn’t the only one that is ramping up its investments into AI-enabled data centers. Rival cloud service providers are all investing in either upgrading or opening new data centers to capture a larger chunk of business from developers and users of large language models (LLMs). In a report published in October 2024, Bloomberg Intelligence estimated that demand for generative AI would push Microsoft, AWS, Google, Oracle, Meta, and Apple would between them devote $200 billion to capex in 2025, up from $110 billion in 2023. Microsoft is one of the biggest spenders, followed closely by Google and AWS, Bloomberg Intelligence said. Its estimate of Microsoft’s capital spending on AI, at $62.4 billion for calendar 2025, is lower than Smith’s claim that the company will invest $80 billion in the fiscal year to June 30, 2025. Both figures, though, are way higher than Microsoft’s 2020 capital expenditure of “just” $17.6 billion. The majority of the increased spending is tied to cloud services and the expansion of AI infrastructure needed to provide compute capacity for OpenAI workloads. Separately, last October Amazon CEO Andy Jassy said his company planned total capex spend of $75 billion in 2024 and even more in 2025, with much of it going to AWS, its cloud computing division.

John Deere unveils more autonomous farm machines to address skill labor shortage

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More Self-driving tractors might be the path to self-driving cars. John Deere has revealed a new line of autonomous machines and tech across agriculture, construction and commercial landscaping. The Moline, Illinois-based John Deere has been in business for 187 years, yet it’s been a regular as a non-tech company showing off technology at the big tech trade show in Las Vegas and is back at CES 2025 with more autonomous tractors and other vehicles. This is not something we usually cover, but John Deere has a lot of data that is interesting in the big picture of tech. The message from the company is that there aren’t enough skilled farm laborers to do the work that its customers need. It’s been a challenge for most of the last two decades, said Jahmy Hindman, CTO at John Deere, in a briefing. Much of the tech will come this fall and after that. He noted that the average farmer in the U.S. is over 58 and works 12 to 18 hours a day to grow food for us. And he said the American Farm Bureau Federation estimates there are roughly 2.4 million farm jobs that need to be filled annually; and the agricultural work force continues to shrink. (This is my hint to the anti-immigration crowd). John Deere’s autonomous 9RX Tractor. Farmers can oversee it using an app. While each of these industries experiences their own set of challenges, a commonality across all is skilled labor availability. In construction, about 80% percent of contractors struggle to find skilled labor. And in commercial landscaping, 86% of landscaping business owners can’t find labor to fill open positions, he said. “They have to figure out how to do

2025 playbook for enterprise AI success, from agents to evals

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More 2025 is poised to be a pivotal year for enterprise AI. The past year has seen rapid innovation, and this year will see the same. This has made it more critical than ever to revisit your AI strategy to stay competitive and create value for your customers. From scaling AI agents to optimizing costs, here are the five critical areas enterprises should prioritize for their AI strategy this year. 1. Agents: the next generation of automation AI agents are no longer theoretical. In 2025, they’re indispensable tools for enterprises looking to streamline operations and enhance customer interactions. Unlike traditional software, agents powered by large language models (LLMs) can make nuanced decisions, navigate complex multi-step tasks, and integrate seamlessly with tools and APIs. At the start of 2024, agents were not ready for prime time, making frustrating mistakes like hallucinating URLs. They started getting better as frontier large language models themselves improved. “Let me put it this way,” said Sam Witteveen, cofounder of Red Dragon, a company that develops agents for companies, and that recently reviewed the 48 agents it built last year. “Interestingly, the ones that we built at the start of the year, a lot of those worked way better at the end of the year just because the models got better.” Witteveen shared this in the video podcast we filmed to discuss these five big trends in detail. Models are getting better and hallucinating less, and they’re also being trained to do agentic tasks. Another feature that the model providers are researching is a way to use the LLM as a judge, and as models get cheaper (something we’ll cover below), companies can use three or more models to

OpenAI’s red teaming innovations define new essentials for security leaders in the AI era

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More OpenAI has taken a more aggressive approach to red teaming than its AI competitors, demonstrating its security teams’ advanced capabilities in two areas: multi-step reinforcement and external red teaming. OpenAI recently released two papers that set a new competitive standard for improving the quality, reliability and safety of AI models in these two techniques and more. The first paper, “OpenAI’s Approach to External Red Teaming for AI Models and Systems,” reports that specialized teams outside the company have proven effective in uncovering vulnerabilities that might otherwise have made it into a released model because in-house testing techniques may have missed them. In the second paper, “Diverse and Effective Red Teaming with Auto-Generated Rewards and Multi-Step Reinforcement Learning,” OpenAI introduces an automated framework that relies on iterative reinforcement learning to generate a broad spectrum of novel, wide-ranging attacks. Going all-in on red teaming pays practical, competitive dividends It’s encouraging to see competitive intensity in red teaming growing among AI companies. When Anthropic released its AI red team guidelines in June of last year, it joined AI providers including Google, Microsoft, Nvidia, OpenAI, and even the U.S.’s National Institute of Standards and Technology (NIST), which all had released red teaming frameworks. Investing heavily in red teaming yields tangible benefits for security leaders in any organization. OpenAI’s paper on external red teaming provides a detailed analysis of how the company strives to create specialized external teams that include cybersecurity and subject matter experts. The goal is to see if knowledgeable external teams can defeat models’ security perimeters and find gaps in their security, biases and controls that prompt-based testing couldn’t find. What makes OpenAI’s recent papers noteworthy is how well they define using human-in-the-middle

Stay Ahead, Stay ONMINE