Anthropic’s chief scientist on 5 ways agents will be even better in 2025

Stay Ahead, Stay ONMINE

Anthropic’s chief scientist on 5 ways agents will be even better in 2025

Agents are the hottest thing in tech right now. Top firms from Google DeepMind to OpenAI to Anthropic are racing to augment large language models with the ability to carry out tasks by themselves. Known as agentic AI in industry jargon, such systems have fast become the new target of Silicon Valley buzz. Everyone from Nvidia to Salesforce is talking about how they are going to upend the industry. “We believe that, in 2025, we may see the first AI agents ‘join the workforce’ and materially change the output of companies,” Sam Altman claimed in a blog post last week. In the broadest sense, an agent is a software system that goes off and does something, often with minimal to zero supervision. The more complex that thing is, the smarter the agent needs to be. For many, large language models are now smart enough to power agents that can do a whole range of useful tasks for us, such as filling out forms, looking up a recipe and adding the ingredients to an online grocery basket, or using a search engine to do last-minute research before a meeting and producing a quick bullet-point summary. In October, Anthropic showed off one of the most advanced agents yet: an extension of its Claude large language model called computer use. As the name suggests, it lets you direct Claude to use a computer much as a person would, by moving a cursor, clicking buttons, and typing text. Instead of simply having a conversation with Claude, you can now ask it to carry out on-screen tasks for you. Anthropic notes that the feature is still cumbersome and error-prone. But it is already available to a handful of testers, including third-party developers at companies such as DoorDash, Canva, and Asana. Computer use is a glimpse of what’s to come for agents. To learn what’s coming next, MIT Technology Review talked to Anthropic’s cofounder and chief scientist Jared Kaplan. Here are five ways that agents are going to get even better in 2025. (Kaplan’s answers have been lightly edited for length and clarity.) 1/ Agents will get better at using tools “I think there are two axes for thinking about what AI is capable of. One is a question of how complex the task is that a system can do. And as AI systems get smarter, they’re getting better in that direction. But another direction that’s very relevant is what kinds of environments or tools the AI can use. “So, like, if you go back almost 10 years now to [DeepMind’s Go-playing model] AlphaGo, we had AI systems that were superhuman in terms of how well they could play board games. But if all you can work with is a board game, then that’s a very restrictive environment. It’s not actually useful, even if it’s very smart. With text models, and then multimodal models, and now computer use—and perhaps in the future with robotics—you’re moving toward bringing AI into different situations and tasks, and making it useful. “We were excited about computer use basically for that reason. Until recently, with large language models, it’s been necessary to give them a very specific prompt, give them very specific tools, and then they’re restricted to a specific kind of environment. What I see is that computer use will probably improve quickly in terms of how well models can do different tasks and more complex tasks. And also to realize when they’ve made mistakes, or realize when there’s a high-stakes question and it needs to ask the user for feedback.” 2/ Agents will understand context “Claude needs to learn enough about your particular situation and the constraints that you operate under to be useful. Things like what particular role you’re in, what styles of writing or what needs you and your organization have. “I think that we’ll see improvements there where Claude will be able to search through things like your documents, your Slack, etc., and really learn what’s useful for you. That’s underemphasized a bit with agents. It’s necessary for systems to be not only useful but also safe, doing what you expected. “Another thing is that a lot of tasks won’t require Claude to do much reasoning. You don’t need to sit and think for hours before opening Google Docs or something. And so I think that a lot of what we’ll see is not just more reasoning but the application of reasoning when it’s really useful and important, but also not wasting time when it’s not necessary.” 3/ Agents will make coding assistants better “We wanted to get a very initial beta of computer use out to developers to get feedback while the system was relatively primitive. But as these systems get better, they might be more widely used and really collaborate with you on different activities. “I think DoorDash, the Browser Company, and Canva are all experimenting with, like, different kinds of browser interactions and designing them with the help of AI. “My expectation is that we’ll also see further improvements to coding assistants. That’s something that’s been very exciting for developers. There’s just a ton of interest in using Claude 3.5 for coding, where it’s not just autocomplete like it was a couple of years ago. It’s really understanding what’s wrong with code, debugging it—running the code, seeing what happens, and fixing it.” 4/ Agents will need to be made safe “We founded Anthropic because we expected AI to progress very quickly and [thought] that, inevitably, safety concerns were going to be relevant. And I think that’s just going to become more and more visceral this year, because I think these agents are going to become more and more integrated into the work we do. We need to be ready for the challenges, like prompt injection. [Prompt injection is an attack in which a malicious prompt is passed to a large language model in ways that its developers did not foresee or intend. One way to do this is to add the prompt to websites that models might visit.] “Prompt injection is probably one of the No.1 things we’re thinking about in terms of, like, broader usage of agents. I think it’s especially important for computer use, and it’s something we’re working on very actively, because if computer use is deployed at large scale, then there could be, like, pernicious websites or something that try to convince Claude to do something that it shouldn’t do. “And with more advanced models, there’s just more risk. We have a robust scaling policy where, as AI systems become sufficiently capable, we feel like we need to be able to really prevent them from being misused. For example, if they could help terrorists—that kind of thing. “So I’m really excited about how AI will be useful—it’s actually also accelerating us a lot internally at Anthropic, with people using Claude in all kinds of ways, especially with coding. But, yeah, there’ll be a lot of challenges as well. It’ll be an interesting year.”

“We believe that, in 2025, we may see the first AI agents ‘join the workforce’ and materially change the output of companies,” Sam Altman claimed in a blog post last week.

In the broadest sense, an agent is a software system that goes off and does something, often with minimal to zero supervision. The more complex that thing is, the smarter the agent needs to be. For many, large language models are now smart enough to power agents that can do a whole range of useful tasks for us, such as filling out forms, looking up a recipe and adding the ingredients to an online grocery basket, or using a search engine to do last-minute research before a meeting and producing a quick bullet-point summary.

In October, Anthropic showed off one of the most advanced agents yet: an extension of its Claude large language model called computer use. As the name suggests, it lets you direct Claude to use a computer much as a person would, by moving a cursor, clicking buttons, and typing text. Instead of simply having a conversation with Claude, you can now ask it to carry out on-screen tasks for you.

Anthropic notes that the feature is still cumbersome and error-prone. But it is already available to a handful of testers, including third-party developers at companies such as DoorDash, Canva, and Asana.

Computer use is a glimpse of what’s to come for agents. To learn what’s coming next, MIT Technology Review talked to Anthropic’s cofounder and chief scientist Jared Kaplan. Here are five ways that agents are going to get even better in 2025.

(Kaplan’s answers have been lightly edited for length and clarity.)

1/ Agents will get better at using tools

“I think there are two axes for thinking about what AI is capable of. One is a question of how complex the task is that a system can do. And as AI systems get smarter, they’re getting better in that direction. But another direction that’s very relevant is what kinds of environments or tools the AI can use.

“So, like, if you go back almost 10 years now to [DeepMind’s Go-playing model] AlphaGo, we had AI systems that were superhuman in terms of how well they could play board games. But if all you can work with is a board game, then that’s a very restrictive environment. It’s not actually useful, even if it’s very smart. With text models, and then multimodal models, and now computer use—and perhaps in the future with robotics—you’re moving toward bringing AI into different situations and tasks, and making it useful.

“We were excited about computer use basically for that reason. Until recently, with large language models, it’s been necessary to give them a very specific prompt, give them very specific tools, and then they’re restricted to a specific kind of environment. What I see is that computer use will probably improve quickly in terms of how well models can do different tasks and more complex tasks. And also to realize when they’ve made mistakes, or realize when there’s a high-stakes question and it needs to ask the user for feedback.”

2/ Agents will understand context

“Claude needs to learn enough about your particular situation and the constraints that you operate under to be useful. Things like what particular role you’re in, what styles of writing or what needs you and your organization have.

“I think that we’ll see improvements there where Claude will be able to search through things like your documents, your Slack, etc., and really learn what’s useful for you. That’s underemphasized a bit with agents. It’s necessary for systems to be not only useful but also safe, doing what you expected.

“Another thing is that a lot of tasks won’t require Claude to do much reasoning. You don’t need to sit and think for hours before opening Google Docs or something. And so I think that a lot of what we’ll see is not just more reasoning but the application of reasoning when it’s really useful and important, but also not wasting time when it’s not necessary.”

3/ Agents will make coding assistants better

“We wanted to get a very initial beta of computer use out to developers to get feedback while the system was relatively primitive. But as these systems get better, they might be more widely used and really collaborate with you on different activities.

“I think DoorDash, the Browser Company, and Canva are all experimenting with, like, different kinds of browser interactions and designing them with the help of AI.

“My expectation is that we’ll also see further improvements to coding assistants. That’s something that’s been very exciting for developers. There’s just a ton of interest in using Claude 3.5 for coding, where it’s not just autocomplete like it was a couple of years ago. It’s really understanding what’s wrong with code, debugging it—running the code, seeing what happens, and fixing it.”

4/ Agents will need to be made safe

“We founded Anthropic because we expected AI to progress very quickly and [thought] that, inevitably, safety concerns were going to be relevant. And I think that’s just going to become more and more visceral this year, because I think these agents are going to become more and more integrated into the work we do. We need to be ready for the challenges, like prompt injection.

[Prompt injection is an attack in which a malicious prompt is passed to a large language model in ways that its developers did not foresee or intend. One way to do this is to add the prompt to websites that models might visit.]

“Prompt injection is probably one of the No.1 things we’re thinking about in terms of, like, broader usage of agents. I think it’s especially important for computer use, and it’s something we’re working on very actively, because if computer use is deployed at large scale, then there could be, like, pernicious websites or something that try to convince Claude to do something that it shouldn’t do.

“And with more advanced models, there’s just more risk. We have a robust scaling policy where, as AI systems become sufficiently capable, we feel like we need to be able to really prevent them from being misused. For example, if they could help terrorists—that kind of thing.

“So I’m really excited about how AI will be useful—it’s actually also accelerating us a lot internally at Anthropic, with people using Claude in all kinds of ways, especially with coding. But, yeah, there’ll be a lot of challenges as well. It’ll be an interesting year.”

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.

IBM Research: When AI and quantum merge

IBM’s research laboratory in Zurich. A look inside an IBM Quantum System Two. Advances in tape development. On the left is a quantum-secure tape drive. Scanning tunneling microscope in one of the Zurich laboratories. The innovation won the Nobel Prize in Physics in 1986.

Petronas launches Malaysia Bid Round 2026

Intel says Google engineers spotted Xeon vulnerabilities

“In a perfect world, the [Trusted Computer Base] would be bug-free; in reality, the complexity of modern systems makes continuous assessment essential. Collaborative reviews allow industry leaders to proactively fix vulnerabilities while fostering transparency for everyone who relies on the technology,” Google researchers wrote. The main problem arose when using

How Cisco’s platform mindset is meeting the AI era

4. Sovereignty, trust, and the rise of sovereign AI In EMEA, trust and sovereignty were more than talk—they were central to almost every discussion. This came across loud and clear at the event and in Davos in January. Cisco emphasized four dimensions of trust: security, innovation, execution, and sovereign control.

Market Positioning Itself for Period of Higher Geopolitical Risk

The oil market is positioning itself for a period of higher geopolitical risk and unpredictability, according to Emily Ashford, Energy Research Head at Standard Chartered Bank. “This is evident in both the gradual push higher in flat price, adjustments in the forward curve and volatility skew,” Ashford said in a report sent to Rigzone by the Standard Chartered team this week. “Brent crude for April delivery settled at $69.04 per barrel on 9 February, a week on week rise of $2.74 per barrel (or 4.13 percent). The forward curve has moved higher week on week,” Ashford added. “While movement at the front of the curve is the most significant, even delivery in five years is $0.53 per barrel higher week on week,” Ashford continued. “Call skew is very well bid, with 1m risk reversal (RR) at 18.69 on 9 February; approximately the same levels as during the initial Russian invasion of Ukraine (19.14 on 8 March 2022) , and was only recently higher in mid-June 2025 (23.01 on 17 June 2025), during the previous escalation in the U.S.-Iran-Israel conflict,” the Standard Chartered Bank Energy Research Head went on to state. In the report, Ashford highlighted that traders and analysts were meeting in London this week for the annual International Energy Week, which the energy head said “is always a useful gauge of market sentiment”. “Last year the focus was around the new second Trump administration, and whether it would be inherently bullish or bearish for oil prices,” Ashford noted. “This was particularly so given the signing of a National Security Presidential Memorandum in early February 2025 that restored maximum pressure on Iran, with a focus on reducing its crude exports,” Ashford said. “One year later, and the complex U.S.-Iran relationship and potential impact on Iranian barrels in the market remains

Secretary Wright Delivers Remarks Alongside Interim Venezuelan President Delcy Rodriguez

CARACAS, VENEZUELA—U.S. Secretary of Energy Chris Wright deliveredremarks today alongside Interim President Delcy Rodriguez after meetings at the Miraflores Palace in Caracas. Secretary Wright’s full remarks are below: Thank you so much, Interim President Rodriguez. It is an honor to stand here with you today, and to be among the tremendous people of Venezuela. As Interim President Rodriguez mentioned, our countries share a long history. It has gone through different chapters as all relationships do. But today, I bring a message from President Trump: He is passionately committed to absolutely transforming the relationship between the United States and Venezuela. This is part of a broader agenda to make the Americas great again, to bring our countries closer together, and to bring commerce, peace, prosperity, jobs, and opportunity to the people of Venezuela in partnership with the United States. These are not just words or ambitions. We have very specific plans and very specific actions already. This is President Trump’s broader agenda: peace, commerce, and trade, not conflict, not military action, not what has dominated so much of our world. Whether it’s in the Middle East, whether it’s in South Asia, or maybe, most importantly, in the Americas, we want commerce, we want peace, we want prosperity, we want security. Our government in Washington, DC, has been working 7 days a week to issue licenses, so existing businesses in Venezuela, new businesses that want to enter Venezuela, Venezuelan national companies can buy products, invest money, raise oil production, create new jobs, grow export revenue. All of the things that have constricted the Venezuelan economy, we want to set the Venezuelan people and the economy free. We had very wonderful and candid dialogues today. We both spoke very candidly about the tremendous opportunities in front of us, and some of the problems

Energy Department Announces 26 Genesis Mission Science and Technology Challenges to Accelerate AI-Enabled American Innovation and Leadership

WASHINGTON—The U.S. Department of Energy (DOE) today announced 26 science and technology challenges of national importance to advance the Genesis Mission and accelerate innovation and discovery through artificial intelligence (AI). Building on President Trump’s Executive Orders Launching The Genesis Mission and Removing Barriers to American Leadership In Artificial Intelligence, the challenges span DOE’s discovery science, energy, and national security missions. Each was selected for its potential to deliver measurable benefits for the American people and to accelerate advancements through the Genesis Mission’s AI platforms, world-class facilities, and public-private partnerships. “These challenges represent a bold step toward a future where science moves at the speed of imagination because of AI. It’s a game-changer for science, energy, and national security,” said DOE Under Secretary for Science and Genesis Mission Lead Dr. Darío Gil. “By uniting the U.S. Government’s unparalleled data resources and DOE’s experimental facilities with cutting-edge AI, we can unlock discoveries that will power the economy, secure our energy future, and keep America at the forefront of global innovation.” “President Trump’s Genesis Mission is mobilizing America’s unmatched scientific infrastructure and AI ingenuity to double the pace of discovery. These 26 challenges are a direct call to action to America’s researchers and innovators to join the Genesis Mission and deliver science and technology breakthroughs that will benefit the American people,” said Assistant to the President and Director of The White House Office of Science and Technology Policy Michael Kratsios. “We look forward to expanding the list of challenges across Federal agencies to bring even greater impact to the Mission.” Working in partnership with DOE’s National Laboratories, industry, and academia, these efforts will deliver tangible results for the American people. Examples include: Scaling the Grid to Power the American Economy: Using AI to improve power grid planning, interconnection, operations, and security — enabling decisions up to 20–100 times faster and improving electricity cost and reliability by up to 10 percent. Harnessing America’s Historic Nuclear Data: Digitizing eight decades of

Oil Drops Amid Supply And Equity Jitters

Oil slid as risk-averse sentiment pervaded global markets and investors digested fresh developments in US-Iran tensions that continue to cloud the supply outlook. West Texas Intermediate fell 2.8% to end the day below $63 a barrel, as equities weakened amid concerns over technology profits. US President Donald Trump said negotiations with Iran over its nuclear program could stretch for as long as a month, but it would be “very traumatic” for Tehran if it failed to reach an agreement. Traders remain concerned about the potential for military strikes and risks to supply from the Middle East. Crude has gained every week this year, with a single exception, as geopolitical risks and supply disruptions drove futures higher. The US intervened in Venezuela in January, then pivoted to Iran after a wave of protests challenged the Islamic Republic’s leadership. A vast hoard of sanctioned oil at sea is also keeping a floor under prices as buyers compete for other barrels. Still, banks maintain that there is abundant supply, with Goldman Sachs Group Inc. saying that the surplus was appearing, but mainly in locations that are less significant for price-setting. The International Energy Agency said on Thursday that global oil stockpiles grew at the strongest pace since 2020 last year, underscoring the view that a period of oversupply has arrived, even if it is not being felt evenly across the globe. As the US-Iranian tensions play out, the Pentagon has positioned a naval force in the region. Israeli President Benjamin Netanyahu met with Trump on Wednesday, with the US leader saying his “preference” is to reach a deal with Tehran on its nuclear ambitions. Netanyahu visited Washington in a bid to warn Trump against such a move, instead hoping to press him to endorse a more sweeping rollback of Iran’s military influence in

Verde Clean Fuels suspends Permian basin GTG project

@import url(‘https://fonts.googleapis.com/css2?family=Inter:[email protected]&display=swap’); a { color: var(–color-primary-main); } .ebm-page__main h1, .ebm-page__main h2, .ebm-page__main h3, .ebm-page__main h4, .ebm-page__main h5, .ebm-page__main h6 { font-family: Inter; } body { line-height: 150%; letter-spacing: 0.025em; font-family: Inter; } button, .ebm-button-wrapper { font-family: Inter; } .label-style { text-transform: uppercase; color: var(–color-grey); font-weight: 600; font-size: 0.75rem; } .caption-style { font-size: 0.75rem; opacity: .6; } #onetrust-pc-sdk [id*=btn-handler], #onetrust-pc-sdk [class*=btn-handler] { background-color: #c19a06 !important; border-color: #c19a06 !important; } #onetrust-policy a, #onetrust-pc-sdk a, #ot-pc-content a { color: #c19a06 !important; } #onetrust-consent-sdk #onetrust-pc-sdk .ot-active-menu { border-color: #c19a06 !important; } #onetrust-consent-sdk #onetrust-accept-btn-handler, #onetrust-banner-sdk #onetrust-reject-all-handler, #onetrust-consent-sdk #onetrust-pc-btn-handler.cookie-setting-link { background-color: #c19a06 !important; border-color: #c19a06 !important; } #onetrust-consent-sdk .onetrust-pc-btn-handler { color: #c19a06 !important; border-color: #c19a06 !important; } <!–> Verde Clean Fuels Inc., Houston, has suspended development of a natural gas-to-gasoline (GTG) project in the Permian basin. The company cited “changing market conditions driven by increasing demand for natural gas,” in the region. ]–> <!–> –><!–> –> June 5, 2024 In 2024, Verde Clean Fuels and Midland-based Diamondback Energy Inc. subsidiary Cottonmouth Ventures LLC agreed to a joint development of a GTG plant in Martin County, Tex., in the Permian’s Midland basin. The project was originally slated to combine Verde’s technology with a feedstock of stranded or otherside-flared associated natural from Diamondback’s Permian basin operations for commercial-scale production of almost 3,000 b/d of fully finished reformulated blend stock for oxygenate blending (RBOB) gasoline. A front-end engineering and design (FEED) study was completed in December 2025. Verde’s chief executive officer, Ernest Miller, said knowledge collected from the work completed “will continue to be useful as we explore other opportunities to deploy our technology.” He said the company will devote resources toward other opportunities “in regions where natural gas is stranded or flared without access to a higher value outlet to market.”

Oil prices ride the Iran seesaw

Oil, fundamental analysis Crude prices were up-and-down this week as traders eyed both optimistic and pessimistic signs regarding the proposed US/Iran talks held Friday. A large drop in crude and distillate inventories were the main fundamental factors noticed. Despite the rollercoaster ride, US prices remained above the key $60/bbl level. WTI had a High of $65.55/bbl on Wednesday with a weekly Low of $61.10 on Tuesday. Brent crude’s High was $69.75/bbl on Wednesday while its Low was $65.75 Monday. Both grades settled lower week-on-week. The WTI/Brent spread has tightened to ($4.50). The week started with a bearish tone as US President Trump spoke with optimism about the upcoming US/Iran talks. However, later in the week, the Iranian government objected to the specific topics the US wants to discuss beyond their nuclear developments and the meeting appeared doomed. The two sides did decide to follow through with the planned meeting Friday with Iranian officials labeling the meeting as a ‘very good start.’ Meanwhile, the Iranian Revolutionary Guard Corps (IRGCC) attempted to stop a US-flagged vessel in the Strait of Hormuz and sent drones to a US Navy ship in the area. Both attempts were thwarted by US naval forces. Some Very Large Crude Carriers (VLCC) are said to have been increasing their normal speeds for faster passage through the Strait of Hormuz while it’s still open. It is estimate that slightly more than 25% of global oil supplies move through the Strait. Russia is now relying heavily on its relationship with China to buy its oil exports after the US offered India a deal whereby tariffs could be cut if India halts buying Russian Urals. Additionally, Indian refiners would have access to Venezuelan exports. No official action has yet to be taken by the Indian government in terms of such a

Data center capex to hit $1.7 trillion by 2030 due to AI boom

Asked on Thursday about the fact that capex is expected to approach the $1 trillion mark in 2026, he said it is somewhat surprising. “Last year, I thought it would take at least three years to get to that trillion dollar mark,” he said. “It seems increases are supported by the result of larger models needed for training infrastructure, and in turn, you need inference as well. You also need a supporting infrastructure in storage, networking, power, and cooling.” AI, he said, has become “the tide that lift all boats, meaning that in addition to the core accelerated compute, AI also positively impacts complementary infrastructure, such as storage, networking, and physical infrastructure.” Fung added that while much of the achievement of projected spending estimates will depend on whether or not this growth is sustainable, he pointed out, “it seems like the large hyperscalers have a lot of weight in optimizing cash flow and cost structures. They’re trying to get as creative as possible, generally moving towards a more vertical, integrated stack with their own custom networking and external financing, which would help [create] more sustainable deployments and operations.” Enterprises thinking of expanding their own infrastructure can learn from this growth. In a recent article on the hyper spending of hyperscalers, Greyhound Research chief analyst Sanchit Vir Gogia said their capex spending levels can help pinpoint where the hyperscalers are expecting bottlenecks, which is useful information for enterprises planning their own cloud strategy across multiple geographies. These and other factors can help enterprises plan their own execution timelines, he said. This article originally appeared on CIO.com.

Cisco highlights memory costs, Silicon One growth in Q2 recap

“AI infrastructure orders taken from hyperscalers totaled $2.1 billion in Q2 compared to $1.3 billion just last quarter and equal to the total orders taken in all of fiscal year ’25, marking another significant acceleration in growth across our silicon, systems and optics,” Robbins said. “Given the strong demand for our Silicon One systems and optics, we now expect to take AI orders in excess of $5 billion and to recognize over $3 billion in AI infrastructure revenue from hyperscalers in FY ’26.” Regarding enterprise uptake, Robbins said Cisco took in $350 million in AI orders from enterprise customers in Q2 and has a pipeline in excess of $2.5 billion for its high-performance AI infrastructure portfolio. Cisco is seeing early enterprise use cases for AI around fraud detection and video analytics in sectors such as financial, manufacturing and pharmaceuticals, for example. “I also see examples in retail, where customers are leveraging agents on mobile devices in retail to help their staff do a better job engaging with their customers. We’re seeing a combination of both investment in cloud-based architectures as well as on prem,” Robbins said. Networking rules Cisco is experiencing a faster-than-historical ramp-up of next-generation platforms, including its Catalyst 9K, Wi-Fi 7, and smart switches, stated Sebastien Naji, a research analyst with William Blair, in a report after the call. He attributed it to three factors: an accelerated refresh cycle in the data center; early AI-readiness efforts in the enterprise; and end-of-support for legacy Catalyst and Nexus switches. “We are seeing strong demand for our next-generation switching, routing and wireless products, which continue to ramp faster than prior product launches. We’re delivering AI-native capabilities across these products, including weaving security into the fabric of the network and modernizing the operational stack of campus networks,” Robbins said. Co-packaged optics? When asked

Energy providers seek flexible load strategies for data center operations

“In theory, yes, they’d have to wait a little bit longer while their queries are routed to a data center that has capacity,” said Lawrence. The one thing the industry cannot do is operate like it has in the past, where data center power was tuned and then forgotten for six months. Previously, data centers would test their power sources once or twice a year. They don’t have that luxury anymore. They need to check their power sources and loads far more regularly, according to Lawrence. “I think that for that for the data center industry to continue to survive like we all need it, there’s going to have to be some realignment on the incentives to why somebody would become flexible,” said Lawrence. The survey suggests that utilities and load operators expect to expand their demand response activities and budgets in the near term. Sixty-three percent of respondents anticipate DR program funding to grow by 50% or more over the next three years. While they remain a major source of load growth and system strain, 57% of respondents indicate that onsite power generation from data centers will be most important to improving grid stability over the next five years. One of the proposed fixes to the power shortage has been small modular nuclear reactors. These have gained a lot of traction in the marketplace even if they have nothing to sell yet. But Lawrence said that that’s not an ideal solution for existing power generators, ironically enough.

Nokia predicts huge WAN traffic growth, but experts question assumptions

Consumer, which includes both mobile access and fixed access, including fixed wireless access. Enterprise and industrial, which covers wide-area connectivity that supports knowledge work, automation, machine vision, robotics coordination, field support, and industrial IoT. AI, including applications that people directly invoke, such as assistants, copilots, and media generation, as well as autonomous use cases in which AI systems trigger other AI systems to perform functions and move data across networks. The report outlines three scenarios: conservative, moderate, and aggressive. “Our goal is to present scenarios that fall within a realistic range of possible outcomes, encouraging stakeholders to plan across the full spectrum of high-impact demand possibilities,” the report says. Nokia’s prediction for global WAN traffic growth ranges from a 13% CAGR for the conservative scenario to 16% CAGR for moderate and 22% CAGR for aggressive. Looking more closely at the moderate scenario, it’s clear that consumer traffic dominates. Enterprise and industrial traffic make up only about 14% to 17% of overall WAN traffic, although their share is expected to grow during the 10-year forecast period. “On the consumer side, the vast majority of traffic by volume is video,” says William Webb, CEO of the consulting firm Commcisive. Asked whether any of that consumer traffic is at some point served up by enterprises, the answer is a decisive “no.” It’s mostly YouTube and streaming services like Netflix, he says. In short, that doesn’t raise enterprise concerns. Nokia predicts AI traffic boom AI is a different story. “Consumer- and enterprise-generated AI traffic imposes a substantial impact on the wide-area network (WAN) by adding AI workloads processed by data centers across the WAN. AI traffic does not stay inside one data center; it moves across edge, metro, core, and cloud infrastructure, driving dense lateral flows and new capacity demands,” the report says. An

Cisco amps up Silicon One line, delivers new systems and optics for AI networking

Those building blocks include the new G300 as well as the G200 51.2 Tbps chip, which is aimed at spine and aggregation applications, and the G100 25.6 Tbps chip, which is aimed at leaf operations. Expanded portfolio of Silicon One P200-powered systems Cisco in October rolled out the P200 Silicon One chip and the high-end, 51.2 Tbps 8223 router aimed at distributed AI workloads. That system supports Octal Small Form-Factor Pluggable (OSFP) and Quad Small Form-Factor Pluggable Double Density (QSFP-DD) optical form factors that help the box support geographically dispersed AI clusters. Cisco grew the G200 family this week with the addition of the 8122X-64EF-O, a 64x800G switch that will run the SONiC OS and includes support for Cisco 800G Linear Pluggable Optics (LPO) connectivity. LPO components typically set up direct links between fiber optic modules, eliminating the need for traditional components such as a digital signal processor. Cisco said its P200 systems running IOS XR software now better support core routing services to allow data-center-to-data-center links and data center interconnect applications. In addition, Cisco introduced a P200-powered 88-LC2-36EF-M line card, which delivers 28.8T of capacity. “Available for both our 8-slot and 18-slot modular systems, this line card enables up to an unprecedented 518.4T of total system bandwidth, the highest in the industry,” wrote Guru Shenoy, senior vice president of the Cisco provider connectivity group, in a blog post about the news. “When paired with Cisco 800G ZR/ZR+ coherent pluggable optics, these systems can easily connect sites over 1,000 kilometers apart, providing the high-density performance needed for modern data center interconnects and core routing.”

NetBox Labs ships AI copilot designed for network engineers, not developers

Natural language for network engineers Beevers explained that network operations teams face two fundamental barriers to automation. First, they lack accurate data about their infrastructure. Second, they aren’t software developers and shouldn’t have to become them. “These are not software developers. They are network engineers or IT infrastructure engineers,” Beevers said. “The big realization for us through the copilot journey is they will never be software developers. Let’s stop trying to make them be. Let’s let these computers that are really good at being software developers do that, and let’s let the network engineers or the data center engineers be really good at what they’re really good at.” That vision drove the development of NetBox Copilot’s natural language interface and its capabilities. Grounding AI in infrastructure reality The challenge with deploying AI in network operations is trust. Generic large language models hallucinate, produce inconsistent results, and lack the operational context to make reliable decisions. NetBox Copilot addresses this by grounding the AI agent in NetBox’s comprehensive infrastructure data model. NetBox serves as the system of record for network and infrastructure teams, maintaining a semantic map of devices, connections, IP addressing, rack layouts, power distribution and the relationships between these elements. Copilot has native awareness of this data structure and the context it provides. This enables queries that would be difficult or impossible with traditional interfaces. Network engineers can ask “Which devices are missing IP addresses?” to validate data completeness, “Who changed this prefix last week?” for change tracking and compliance, or “What depends on this switch?” for impact analysis before maintenance windows.

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