Stay Ahead, Stay ONMINE

CockroachDB’s distributed vector indexing tackles the looming AI data explosion enterprises aren’t ready for

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More As the scale of enterprise AI operations continues to grow, having access to data is no longer enough. Enterprises now must have reliable, consistent and accurate access to data. That’s a realm where distributed SQL database […]

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More


As the scale of enterprise AI operations continues to grow, having access to data is no longer enough. Enterprises now must have reliable, consistent and accurate access to data.

That’s a realm where distributed SQL database vendors play a key role, providing a replicated database platform that can be highly resilient and available. The latest update from Cockroach Labs is all about enabling vector search and agentic AI at distributed SQL scale. CockroachDB 25.2 is out today, promising a 41% efficiency gain, an AI-optimized vector index for distributed SQL scale, and core database improvements that improve both operations and security. 

CockroachDB is one of many distributed SQL options in the market today, including Yugabyte, Amazon Aurora dSQL and Google AlloyDBSince its inception a decade ago, the company has aimed to differentiate itself from rivals by being more resilient. In fact, the name ‘cockroach’ comes from the idea that a cockroach is really hard to kill. This idea remains relevant in the AI era.

“Certainly people are interested in AI, but the reasons people chose Cockroach five years ago, two years ago or even this year seems to be pretty consistent, they need this database to survive,” Spencer Kimball co-founder and CEO of Cockroach Labs told VentureBeat. “AI in our context, is AI mixed with the operational capabilities that Cockroach brings…so to the extent that AI is becoming more important, it’s how does my AI survive, it needs to be just as mission critical as the actual metadata.”

The distributed vector indexing problem facing enterprise AI

Vector capable databases, which are used by AI systems for training as well as for Retrieval Augmented Generation (RAG) scenarios, are commonplace in 2025.

Kimball argued that vector databases today work well on single nodes. They tend to struggle on larger deployments with multiple geographically dispersed nodes, which is what distributed SQL is all about. CockroachDB’s approach tackles the complex problem of distributed vector indexing. The company’s new C-SPANN vector index uses the SPANN algorithm, which is based on Microsoft research. This specifically handles billions of vectors across a distributed, disk-based system.

Understanding the technical architecture reveals why this poses such a complex challenge. Vector indexing in CockroachDB isn’t a separate table; it’s an index type applied to columns within existing tables. Without an index, vector similarity searches perform brute-force linear scans through all data. This works fine for small datasets but becomes prohibitively slow as tables grow. 

The Cockroach Labs engineering team had to solve multiple problems simultaneously: uniform efficiency at massive scale, self-balancing indexes and maintaining accuracy while underlying data changes rapidly.

Kimball explained that the C-SPANN algorithm solves this by creating a hierarchy of partitions for vectors in a very high multi-dimensional space. This hierarchical structure enables efficient similarity searches even across billions of vectors.

Security enhancements address AI compliance challenges

AI applications handle increasingly sensitive data. CockroachDB 25.2 introduces enhanced security features, including row-level security and configurable cipher suites. 

These capabilities address regulatory requirements like DORA and NIS2 that many enterprises struggle to meet.

Cockroach Labs’ research shows 79% of technology leaders report being unprepared for new regulations. Meanwhile, 93% cite concerns over the financial impact of outages averaging over $222,000 annually.

“Security is something that is significantly increasing and I think that the big thing about security to realize is that like many things, it’s impacted dramatically by this AI stuff,” Kimball observed. 

Operational big data for agentic AI set to drive massive growth

The coming wave of AI-driven workloads creates what Kimball terms “operational big data”—a fundamentally different challenge from traditional big data analytics. 

While conventional big data focuses on batch processing large datasets for insights, operational big data demands real-time performance at massive scale for mission-critical applications.

“When you really think about the implications of agentic AI, it’s just a lot more activity hitting APIs and ultimately causing throughput requirements for the underlying databases,” Kimball explained.

The distinction matters enormously. Traditional data systems can tolerate latency and eventual consistency because they support analytical workloads. Operational big data powers live applications where milliseconds matter and consistency can’t be compromised.

AI agents drive this shift by operating at machine speed rather than human pace. Current database traffic comes primarily from humans with predictable usage patterns. Kimball emphasized that AI agents will multiply this activity exponentially.

Performance breakthrough targets AI workload economics

Better economics and efficiency are needed to cope with the growing scale of data access.

Cockroach Labs claims that CockroachDB 25.2 provides a 41% efficiency improvement. Two key optimizations in the release that will help improve overall database efficiency are generic query plans and buffered writes. 

Buffered writes solve a particular problem with object-relational mapping (ORM) generated queries that tend to be “chatty.” These read and write data across distributed nodes inefficiently. The buffered writes feature keeps writes in local SQL coordinators. This eliminates unnecessary network round trips.

“What buffered writes do is that they keep all of the writes that you’re planning to do in the local SQL coordinator,” Kimball explained. “So then if you read from something that you’ve just written, it doesn’t have to go back out to the network.”

Generic query plans solve a fundamental inefficiency in high-volume applications. Most enterprise applications use a limited set of transaction types that get executed millions of times with different parameters. Instead of repeatedly replanning identical query structures, CockroachDB now caches and reuses these plans.

Implementing generic query plans in distributed systems presents unique challenges that single-node databases don’t face. CockroachDB must ensure that cached plans remain optimal across geographically distributed nodes with varying latencies.

“In distributed SQL, the generic query plans, they’re kind of a slightly heavier lift, because now you’re talking about a potentially geo-distributed set of nodes with different latencies,” Kimball explained. “You have to be careful with the generic query plan that you don’t use something that’s suboptimal because you’ve sort of conflated like, oh well, this looks the same.”

What this means for enterprises planning AI and data infrastructure

Enterprise data leaders face immediate decisions as agentic AI threatens to overwhelm the current database infrastructure.

The shift from human-driven to AI-driven workloads will create operational big data challenges that many organizations aren’t prepared for. Preparing now for the inevitable growth in data traffic from agentic AI is a strong imperative. For enterprises leading in AI adoption, it makes sense to invest in a distributed database architecture now that can handle both traditional SQL and vector operations at scale. 

CockroachDB 25.2 offers one potential option, raising the performance and efficiency of distributed SQL to meet the data challenges of agentic AI. Fundamentally, it’s about having the technology in place to scale both vector and traditional data retrieval.

Shape
Shape
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.

Shape

3 ways to streamline cloud adoption and cloud security

In today’s cloud-first world, speed and agility are the currency of innovation. Organizations are under pressure to deliver applications faster, more securely, and across increasingly distributed cloud environments. As a result, your developers are being asked to ship code at a rapid pace, often on a weekly or even daily

Read More »

GenAI controls and ZTNA architecture set SSE vendors apart

Critical SSE capabilities Gartner defines SSE as “an offering that secures access to the web, cloud services, and private applications regardless of the location of the user, the device they are using, or where that application is hosted.” “[SSE] provides a range of security capabilities, including adaptive access based on

Read More »

Understanding devices on Linux systems

$ df -hFilesystem Size Used Avail Use% Mounted on/dev/sda3 14G 6.7G 6.2G 52% /devtmpfs 4.0M 0 4.0M 0% /devtmpfs 886M 96K 886M 1% /dev/shmefivarfs 64K 6.0K 53K 11% /sys/firmware/efi/efivarstmpfs 355M 1.7M 353M 1% /runtmpfs 1.0M 0 1.0M 0% /run/credentials/systemd-journald.servicetmpfs 1.0M 0 1.0M 0% /run/credentials/systemd-network-generator.servicetmpfs 1.0M 0 1.0M 0% /run/credentials/systemd-udev-load-credentials.servicetmpfs 1.0M

Read More »

Broadcom’s 102.4 Tbps Tomahawk 6 targets million-XPU AI clusters

The MMU complexity increase reflects the challenges of managing packet buffering, queue scheduling and congestion control at these extreme bandwidths. Traditional approaches to packet switching become increasingly difficult as the numbers of ports, queues and simultaneous flows grow exponentially. The Tomahawk 6 addresses these challenges through several key architectural innovations.

Read More »

TotalEnergies Acquires Low Carbon’s UK Renewables Portfolio

TotalEnergies SE has acquired a pipeline of eight solar projects with a capacity of 350 megawatts (MW) and two battery storage projects with a capacity of 85 MW from Low Carbon Investment Management Ltd. With the solar projects at an advanced development stage, the aim is to have them up and running by 2028. They are expected to generate over 350 GWh of electricity annually, which is roughly equivalent to the needs of around 100,000 UK homes, TotalEnergies said in a media release. “The acquisition of these solar and battery projects located in the south of England will complement our integrated electricity portfolio in the UK, which includes 1.1 GW of gross installed offshore wind, 1.3 GW of gross combined cycle gas turbine, and more than 600 MW of solar projects under development”, said Olivier Jouny, SVP for renewables at TotalEnergies. “We are very pleased to have finalized this agreement with TotalEnergies, who are making an impressive commitment to building renewable infrastructure and sees Low Carbon deliver one of the largest ready-to-build portfolios of solar and battery projects in the UK market. Once built, these projects will make a vital contribution to the Government’s Clean Power 2030 ambition and reinforce Low Carbon’s track record for developing renewable assets of the highest quality, while supporting the build-out of the rest of our renewables pipeline as we continue to scale as an independent power producer”, Roy Bedlow, Chief Executive and Founder at Low Carbon, added. TotalEnergies’ UK renewable portfolio has 1.1 GW of installed capacity from the Seagreen offshore wind farm and 4.5 GW in development, including offshore wind and solar projects. Totalenergies also has a 50 percent stake in a 1.3 GW combined cycle gas turbine project in partnership with EPUKI. To contact the author, email [email protected] What do you think?

Read More »

Sonardyne Secures Monitoring Deal for NEP CCS Project

Sonardyne International Ltd. has won the bid to provide baseline environmental monitoring services for the UK’s first offshore carbon capture and storage (CCS) site. The company said in a media release that it has been contracted by the Northern Endurance Partnership (NEP) of BP plc, Equinor ASA, and TotalEnergies SE.   NEP is the developer of the onshore and offshore infrastructure needed to transport carbon dioxide (CO2) from carbon capture projects across Teesside and the Humber – collectively known as the East Coast Cluster – for storage under the North Sea. Sonardyne said it will implement environmental monitoring through seabed landers at crucial spots above and around the subsurface Endurance site – the saline aquifer situated 145 kilometers (90.1 miles) from the Teesside coast, where captured CO2 will be stored. “NEP is at the forefront of driving the UK’s journey to net zero in securing a permit for carbon capture and storage at the Endurance site. Our collaboration with Sonardyne to monitor the storage site plays a critical role in ensuring carbon is stored safely and securely”, Sarah Buchanan, BP Subsurface Manager, said. Site monitoring will start in the summer of 2026 for two years before CO2 transportation and storage begins. Seabed landers will use the Edge data processing app, power management, and acoustic communications for remote deployment. Each lander will feature the Origin 600 ADCP, Wavefront’s passive sonar array, and various third-party sensors to detect changes in water chemistry, Sonardyne said. Data will be collected wirelessly without retrieving the landers. Monitoring is crucial for ensuring the safe containment of CO2, reassuring stakeholders, and fostering public confidence in CCS initiatives, the company stated. “We are delighted to be working with NEP in delivering cutting-edge marine technology to ensure their safe and successful offshore carbon storage operations”, Stephen Auld, Business Development Manager for Custom

Read More »

Oil Falls as Saudi Arabia Seeks More Major Production Hikes

Oil fell on signs Saudi Arabia wants another major production increase, raising expectations that a glut of crude will form this year.  West Texas Intermediate slid 0.9% to settle below $63 a barrel, paring losses of almost 2% after Bloomberg News reported that the de facto OPEC leader is open to additional significant output hikes in a bid for market share. The kingdom wants the group to add at least 411,000 barrels a day in August and potentially September, ideally as quickly as possible to take advantage of peak summer demand, according to people familiar with the matter. The cartel’s intentions have already been well telegraphed, despite the market’s knee-jerk reaction, said Rebecca Babin, a senior energy trader at CIBC Private Wealth Group.  “Still, it shows the course that OPEC is currently on will likely continue,” she added.  The development helped jolt crude out of a listless trade on mixed demand data. US government figures showed the country’s crude stockpiles fell 4.3 million barrels last week, raising expectations of near-term tightness. Meanwhile, gasoline demand declined.  Oil rose at the start of the week after a decision by OPEC+ to raise production in July was in line with expectations. Saudi Arabia led increases in OPEC oil production last month as the group began its series of accelerated supply additions, according to a Bloomberg survey. Nevertheless, the hike fell short of the full amount the kingdom could have added under the agreements.  Saudi Aramco, meanwhile, cut its oil prices to Asia after OPEC+ continued with its outsized output increases for a third month. The decrease is smaller than a 35-cents-a-barrel reduction anticipated in a survey of refiners and traders. Prices are still down about 12% this year on fears around a looming supply glut, while traders continue to monitor US trade tariffs as President Donald Trump said his

Read More »

Saudi Arabia Cuts Oil Prices for Asia as OPEC+ Adds More Barrels

Saudi Aramco cut the price of its main oil grade to buyers in Asia after OPEC+ continued with its outsized output increases for a third month. The Saudis led the producer group over the weekend in agreeing to raise production by 411,000 barrels a day in July, a third straight month of outsized hikes. In tandem with US President Donald Trump’s trade war, the supply increases have helped drive benchmark oil prices about 12% lower in London since early April.    State producer Aramco will cut the price for Arab Light crude, its flagship grade, by 20 cents a barrel to $1 a barrel more than the regional benchmark for Asian customers, a price list seen by Bloomberg shows. The decrease is nevertheless smaller than a 35-cents-a-barrel reduction anticipated in a survey of refiners and traders. After surprising markets in April with the announcement of a bigger-than-planned output hike, the Saudis have kept pressing for more oil. The step aims to win back market share and takes away some of the advantage of a handful of OPEC+ members who’ve overproduced. Russia, which at times has exceeded its quota, was among those pushing for a more moderate supply hike in July. Refiners have continued to enjoy healthy margins for products as many countries move into a higher summer demand period. Still, crude stockpiles are starting to increase, indicating supply is outweighing demand, and refinery profits slipped at the end of last month.  WHAT DO YOU THINK? Generated by readers, the comments included herein do not reflect the views and opinions of Rigzone. All comments are subject to editorial review. Off-topic, inappropriate or insulting comments will be removed.

Read More »

Amazon, Google exploring all options for meeting growing power needs

Dive Brief: Amazon and Google are trying to remain nimble as their electricity demands increase steeply during a period of political uncertainty and technological evolution for the energy sector, division heads for the hyperscalers said during a Wednesday panel at the American Council on Renewable Energy’s Finance Forum. As Google attempts to meet a goal of operating on 100% carbon-free energy by 2030, the company is making investments in emerging generation sources like geothermal and small modular nuclear reactors in an attempt to accelerate them, said Will Conkling, Google’s head of data center energy, Americas. “But even at Google, we can’t wish things to be true, right? We have to deal with the reality of technology and commercialization curves.” “It’s going to take a mix of everything to really reach the capacity that we need as a nation,” said Jessica Johnson, director of offtake at CleanCapital. “Not just for data centers, but for the entire demand.” Dive Insight: The last five years have been tumultuous for the renewable energy industry, Johnson said, and recently the sector has seen added uncertainty resulting from President Trump’s tariffs as well as a House-passed budget bill that would slash most tax credits included in the Inflation Reduction Act. As a result, she said many energy buyers that don’t have the resources of Amazon or Google have been “shying away” from renewables and wanting to see “what else is out there.” However, Johnson said, fast-deploying generation like solar and storage still has a place when weighed against resources that are slower to bring online, like nuclear. Conkling said he doesn’t see the challenge of responding to demand growth from data centers as one that can be met by a one-size-fits-all solution. He said that when there is growth, “the system has always basically looked at

Read More »

BlackRock Escapes Texas Oil-Boycott List After ESG Retreat

BlackRock Inc. was removed from Texas’ blacklist of companies that boycott fossil fuels, ending a three-year standoff over the environmental policies of the world’s largest asset manager. The move means pension funds and other state-run investment accounts — which manage more than $300 billion of assets — will be allowed to purchase BlackRock shares, invest in its exchange-traded funds and hire the firm for advice and risk management. Inclusion on the list resulted in some Texas entities pulling billions of dollars of assets from the firm. State Comptroller Glenn Hegar said BlackRock had rolled back many of its green-focused initiatives, including exiting the Net Zero Asset Managers initiative and stepping back from the Climate Action 100+, a group devoted to cutting greenhouse gas emissions.  The move marks a win for BlackRock Chief Executive Officer Larry Fink, who has been courting Texas leaders. Last year, he took the stage with Lieutenant Governor Dan Patrick at a summit focused on shoring up the state’s energy grid, and just months ago BlackRock sponsored a table at the Black Tie & Boots Gala, a celebration of conservative politics in Texas. The company is also backing a Dallas-based Texas Stock Exchange. “We appreciate the Comptroller’s resolution of this matter,” a spokesperson for the firm said in a statement. “BlackRock is proud to help millions of Texans retire with dignity and, on behalf of clients, invests over $400 billion in corporations, local governments, energy infrastructure and other private assets throughout the state. These investments support the continued growth of the Texas economy.” Hegar touted some of BlackRock’s steps. He said while they are “unrelated” to the listing decision, the actions “show a real commitment to overall policy changes and a desire to act as a trusted partner in the growth of the Texas economy.” The company was a major target for conservative activists for years, with

Read More »

LiquidStack launches cooling system for high density, high-powered data centers

The CDU is serviceable from the front of the unit, with no rear or end access required, allowing the system to be placed against the wall. The skid-mounted system can come with rail and overhead piping pre-installed or shipped as separate cabinets for on-site assembly. The single-phase system has high-efficiency dual pumps designed to protect critical components from leaks and a centralized design with separate pump and control modules reduce both the number of components and complexity. “AI will keep pushing thermal output to new extremes, and data centers need cooling systems that can be easily deployed, managed, and scaled to match heat rejection demands as they rise,” said Joe Capes, CEO of LiquidStack in a statement. “With up to 10MW of cooling capacity at N, N+1, or N+2, the GigaModular is a platform like no other—we designed it to be the only CDU our customers will ever need. It future-proofs design selections for direct-to-chip liquid cooling without traditional limits or boundaries.”

Read More »

Enterprises face data center power design challenges

” Now, with AI, GPUs need data to do a lot of compute and send that back to another GPU. That connection needs to be close together, and that is what’s pushing the density, the chips are more powerful and so on, but the necessity of everything being close together is what’s driving this big revolution,” he said. That revolution in new architecture is new data center designs. Cordovil said that instead of putting the power shelves within the rack, system administrators are putting a sidecar next to those racks and loading the sidecar with the power system, which serves two to four racks. This allows for more compute per rack and lower latency since the data doesn’t have to travel as far. The problem is that 1 mW racks are uncharted territory and no one knows how to manage the power, which is considerable now. ”There’s no user manual that says, hey, just follow this and everything’s going to be all right. You really need to push the boundaries of understanding how to work. You need to start designing something somehow, so that is a challenge to data center designers,” he said. And this brings up another issue: many corporate data centers have power plugs that are like the ones that you have at home, more or less, so they didn’t need to have an advanced electrician certification. “We’re not playing with that power anymore. You need to be very aware of how to connect something. Some of the technicians are going to need to be certified electricians, which is a skills gap in the market that we see in most markets out there,” said Cordovil. A CompTIA A+ certification will teach you the basics of power, but not the advanced skills needed for these increasingly dense racks. Cordovil

Read More »

HPE Nonstop servers target data center, high-throughput applications

HPE has bumped up the size and speed of its fault-tolerant Nonstop Compute servers. There are two new servers – the 8TB, Intel Xeon-based Nonstop Compute NS9 X5 and Nonstop Compute NS5 X5 – aimed at enterprise customers looking to upgrade their transaction processing network infrastructure or support larger application workloads. Like other HPE Nonstop systems, the two new boxes include compute, software, storage, networking and database resources as well as full-system clustering and HPE’s specialized Nonstop operating system. The flagship NS9 X5 features support for dual-fabric HDR200 InfiniBand interconnect, which effectively doubles the interconnect bandwidth between it and other servers compared to the current NS8 X4, according to an HPE blog detailing the new servers. It supports up to 270 networking ports per NS9 X system, can be clustered with up to 16 other NS9 X5s, and can support 25 GbE network connectivity for modern data center integration and high-throughput applications, according to HPE.

Read More »

AI boom exposes infrastructure gaps: APAC’s data center demand to outstrip supply by 42%

“Investor confidence in data centres is expected to strengthen over the remainder of the decade,” the report said. “Strong demand and solid underlying fundamentals fuelled by AI and cloud services growth will provide a robust foundation for investors to build scale.” Enterprise strategies must evolve With supply constrained and prices rising, CBRE recommended that enterprises rethink data center procurement models. Waiting for optimal sites or price points is no longer viable in many markets. Instead, enterprises should pursue early partnerships with operators that have robust development pipelines and focus on securing power-ready land. Build-to-suit models are becoming more relevant, especially for larger capacity requirements. Smaller enterprise facilities — those under 5MW — may face sustainability challenges in the long term. The report suggested that these could become “less relevant” as companies increasingly turn to specialized colocation and hyperscale providers. Still, traditional workloads will continue to represent up to 50% of total demand through 2030, preserving value in existing facilities for non-AI use cases, the report added. The region’s projected 15 to 25 GW gap is more than a temporary shortage — it signals a structural shift, CBRE said. Enterprises that act early to secure infrastructure, invest in emerging markets, and align with power availability will be best positioned to meet digital transformation goals. “Those that wait may find themselves locked out of the digital infrastructure they need to compete,” the report added.

Read More »

Cisco bolsters DNS security package

The software can block domains associated with phishing, malware, botnets, and other high-risk categories such as cryptomining or new domains that haven’t been reported previously. It can also create custom block and allow lists and offers the ability to pinpoint compromised systems using real-time security activity reports, Brunetto wrote. According to Cisco, many organizations leave DNS resolution to their ISP. “But the growth of direct enterprise internet connections and remote work make DNS optimization for threat defense, privacy, compliance, and performance ever more important,” Cisco stated. “Along with core security hygiene, like a patching program, strong DNS-layer security is the leading cost-effective way to improve security posture. It blocks threats before they even reach your firewall, dramatically reducing the alert pressure your security team manages.” “Unlike other Secure Service Edge (SSE) solutions that have added basic DNS security in a ‘checkbox’ attempt to meet market demand, Cisco Secure Access – DNS Defense embeds strong security into its global network of 50+ DNS data centers,” Brunetto wrote. “Among all SSE solutions, only Cisco’s features a recursive DNS architecture that ensures low-latency, fast DNS resolution, and seamless failover.”

Read More »

HPE Aruba unveils raft of new switches for data center, campus modernization

And in large-scale enterprise environments embracing collapsed-core designs, the switch acts as a high-performance aggregation layer. It consolidates services, simplifies network architecture, and enforces security policies natively, reducing complexity and operational cost, Gray said. In addition, the switch offers the agility and security required at colocation facilities and edge sites. Its integrated Layer 4 stateful security and automation-ready platform enable rapid deployment while maintaining robust control and visibility over distributed infrastructure, Gray said. The CX 10040 significantly expands the capacity it can provide and the roles it can serve for enterprise customers, according to one industry analyst. “From the enterprise side, this expands on the feature set and capabilities of the original 10000, giving customers the ability to run additional services directly in the network,” said Alan Weckel, co-founder and analyst with The 650 Group. “It helps drive a lower TCO and provide a more secure network.”  Aimed as a VMware alternative Gray noted that HPE Aruba is combining its recently announced Morpheus VM Essentials plug-in package, which offers a hypervisor-based package aimed at hybrid cloud virtualization environments, with the CX 10040 to deliver a meaningful alternative to Broadcom’s VMware package. “If customers want to get out of the business of having to buy VM cloud or Cloud Foundation stuff and all of that, they can replace the distributed firewall, microsegmentation and lots of the capabilities found in the old VMware NSX [networking software] and the CX 10k, and Morpheus can easily replace that functionality [such as VM orchestration, automation and policy management],” Gray said. The 650 Group’s Weckel weighed in on the idea of the CX 10040 as a VMware alternative:

Read More »

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.

Read More »

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

Read More »

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

Read More »

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

Read More »