Although it requires thousands of workers to build these facilities, “The reality is data centers just don’t employ that many people,’’ wrote Tom Dotan, a reporter for The Wall Street Journal, wrote on LinkedIn about his related WSJ article.

Fact vs. fiction about full-time data center jobs

Data centers have been touted by politicians and business leaders as a new avenue for employment. When President Trump discussed OpenAI’s Stargate AI venture during a recent press conference, he said that more than 100,000 new jobs would be created “almost immediately.” OpenAI echoed that in a blog post, saying Stargate would “create hundreds of thousands of American jobs.”

However, Dotan cited a one million square foot facility in Abilene, TX that OpenAI is planning to use for its Stargate AI venture that will employ 1,500 people to build it but is projected to employ only 100 people full time. “That’s one-fifth the number of people who will be working in a nearby cheese packing plant that is a fraction of the size,’’ Dotan wrote.

“Data centers are very labor-intensive to build, but not as labor-intensive to operate,” according to Jim Grice, a real estate and project finance attorney who focuses on data centers, as quoted in the WSJ article.

Synergy Research Group Chief Analyst John Dinsdale said in Dotan’s WSJ article that, while data centers can employ more than 1,000 people in the several months or years it takes to build them, it is rare for them to need more than 100-200 once they open.

Regional economic impact of these data centers is limited 

These sentiments are echoed by Good Jobs First, a Washington, D.C.-based nonprofit that tracks the use of state and local economic development subsidies. “Data centers require large amounts of capital investment, but, unlike manufacturing projects, they create few jobs,’’ the organization wrote in a blog post.

In addition, “Besides electricity and water, data centers buy very little from local communities, so their economic impact on a region is limited,’’ Good Jobs First said.

OpenAI’s estimate of the total jobs resulting from Stargate includes ones created indirectly by company and employee spending in communities, according to a company spokeswoman.

The post WSJ Debunks AI Data Centers Jobs Myth appeared first on eWEEK.

Musk aims to raise the bar on AI chatbots

The billionaire announced Grok 3’s launch last week during a video call at the World Government Summit in Dubai, touting the chatbot as having “very powerful reasoning capabilities” and being the “smartest AI on earth.’’

The AI model was trained on a lot of synthetic data and is designed to self-correct by analyzing inconsistencies, Musk said. If Grok 3 has data that is wrong, it will remove it, he added.

“Grok 3 is scary smart … [it] has very powerful reasoning capabilities, so in the tests that we’ve done thus far, Grok 3 is outperforming anything that’s been released, that we’re aware of, so that’s a good sign,” Musk said.

How Grok 3 fits into the competitive AI market

Grok 3’s launch comes on the heels of the January launch of DeepSeek R-1, which took the AI world by storm and sent the Nasdaq into a dramatic decline with tech stocks. The Chinese startup’s two large language models (LLMs) – DeepSeek R-1 and DeepSeek V-3 – can reportedly solve scientific problems at a similar level to ChatGPT.

Perhaps the biggest shock was DeepSeek’s claim that it only spent about $6 million to train its model — much less than OpenAI’s o1. While some tech experts have questioned that amount, governments and AI companies worldwide are striving to develop AI chatbots that are more sophisticated and cost-effective.

Bad blood between Musk and OpenAI cofounder Sam Altman

Musk’s Grok 3 announcement is the latest strike from the world’s richest man, who cofounded OpenAI with Sam Altman as a non-profit in 2015. Following a longstanding feud between the two over the company’s direction, Musk and a consortium of investors offered $97.4 billion to buy the assets of OpenAI’s nonprofit. OpenAI has said it wants to become a for-profit organization to secure the capital it needs to develop the best AI models.

In August 2024, Musk sued OpenAI CEO Sam Altman and others and has asked a U.S. district judge to block OpenAI’s attempt to transition to a for-profit entity. OpenAI said last week that Musk’s bid conflicted with his lawsuit, and then on Feb. 14 rejected his takeover offer.

The post Grok 3 Launches Today: Elon Musk Calls the Chatbot ‘The Smartest AI on Earth’ appeared first on eWEEK.

“The chasm between AI’s breakneck innovation and the systems and frameworks needed to govern its responsibility — is growing,’’ Andrew Wells, NTT Data’s chief data and AI officer, told eWeek. “Somehow, it’s our nature to believe that, as a market matures, risk drops.”

The AI responsibility gap is widening

Wells said he was surprised by “how fractured the C-suite is on this issue: There is no consensus, statistically, on the responsibility-innovation balance.” More than 60 percent of respondents acknowledge the gap is a significant issue, while 81 percent of top-performing organizations place high importance on balancing innovation with responsibility, Wells noted.

The report, The AI Responsibility Gap: Why Leadership Is the Missing Link, also finds that more than 80 percent of respondent leaders cited unclear government regulations as barriers to AI investment and implementation, leading to delayed adoption.

Significantly, security and ethics are taking lower priority than AI ambitions — the report finds that 89 percent of C-suite leaders worry about AI security risks associated with GenAI deployments. Yet only 24 percent of CISOs believe their organizations follow a robust framework to balance AI risk and value creation.

Against this backdrop, 67 percent of executives said their employees lack the skills to work effectively with AI, while 72 percent admitted they do not have an AI policy in place to guide responsible use.

Another major finding is significant concerns regarding sustainability: 75 percent of leaders said AI ambitions conflict with corporate sustainability goals, forcing organizations to rethink energy-intensive AI solutions.

How leaders can close the AI responsibility gap

Without decisive action, organizations risk a future where AI advancements outstrip the governance needed to ensure ethical, secure, and effective AI adoption, the report warns.

“Innovation without a responsibility mandate, whether through stronger regulatory frameworks or decisive leadership, leaves organizations struggling to balance opportunity with risk,’’ the report warned. “Business leaders who strike this balance are better positioned to benefit fully from AI while safeguarding their organization’s future — but strong, focused leadership is required.”

NTT Data recommends leaders:

• Build AI Responsibly: Develop AI, including GenAI, from the ground up and end-to-end, integrating security, compliance, and transparency into development from day one.
• Exceed Legal Standards: Go beyond legal requirements to meet AI ethical and social standards using a systematic approach.
• Upskill Employees: Ensure teams are trained to work alongside AI and understand AI’s risks and opportunities.
• Strengthen Governance: Collaborate with businesses, regulators, and industry leaders to create clearer, actionable AI governance frameworks and establish global AI standards.

Methodology

The report is based on insights from more than 2,300 C-suite leaders and decision-makers across 34 countries in the fall of 2024, NTT Data said.

The post NTT Data Report: Organizations Struggle to Balance Rapid AI Growth with Accountability and Ethics             appeared first on eWEEK.

Top 10 most captivating AI productivity tools

1. Odoo: 17 minutes and 37 seconds was the average visit duration.
2. Databricks: 12 minutes and 10 seconds was the average visit duration on the data intelligence platform.
3. Replit: Nine minutes and 31 seconds was the average visit duration on the web-based integrated development environment.
4. Coda: Eight minutes and 34 seconds was the average visit duration on the collaboration platform.
5. Gamma: Seven minutes and 24 seconds was the average visit duration on the presentation website. 
6. Replicate: Seven minutes and 19 seconds was the average visit duration on the AI model gallery site.
7. ChatGPT: Six minutes and 17 seconds was the average visit duration.
8. Anthropic’s AI assistant Claude: Five minutes and 59 seconds was the average visit duration.
9. Zapier: Four minutes and 52 seconds was the average visit duration on the workflow automation site.
10. VidIQ: Three minutes and 57 seconds was the average visit duration of the video production assistant.

How these AI tools were ranked

Reveal collated 101 AI-powered tools with the most organic traffic and then ranked them by their average visit duration between October and December 2024 to determine which tool users spend the most time on. The firm said the organic traffic data was obtained from marketing software firm Ahrefs, and the average visit duration data was obtained from analytics tool SimilarWeb. The list was then filtered to AI-powered tools focused on productivity.

Why Reveal focused on productivity

The ability to enhance productivity “is one of the best applications for AI since an AI can analyze vast amounts of data in a fraction of the time it would take for a human to do the same, for example,” said Wendell Jisa, founder and CEO of Reveal, in a statement. “This allows people to work more creatively and efficiently, spending more time on tasks humans excel at.”

With that in mind, Reveal wanted to see which AI-powered tools users spent the most time on, Jisa said. “What is particularly impressive is that, on average, Odoo users spent over five minutes longer on the app than the next best app and over 11 minutes more than on ChatGPT,’’ Jisa said. “This could be because Odoo implements many clearly labeled dedicated AI-powered apps on its platform for optimizing and simplifying business operations, while ChatGPT relies more on the user to figure out how and where it can be used effectively.”

The post Odoo Outranks ChatGPT Among the Most Captivating AI Productivity Tools appeared first on eWEEK.

The findings indicate that professionals in computer science (44%) and medicine (38%) fields and those earlier in their careers (39%) are more inclined to adopt AI early. In contrast, those specializing in life sciences and those late in their careers (41%) tend to be more cautious about integrating AI into their work. Roughly two-thirds of mid-career researchers want to be early or average adopters of AI technology.

In addition, AI adoption varies across regions. Researchers in China (59%) and Germany (57%) are leading in integrating AI into their work, compared to the global adoption rate of 44%. Additionally, 57% of early career researchers have already used AI in their work.

The study was based on insights from nearly 5,000 researchers worldwide in 2024, Wiley said.

Barriers to AI adoption and current use

The study identified the lack of clear guidance and training as major barriers to wider AI adoption, with 63% of researchers citing these as challenges. As a result, AI usage remains concentrated on a few core tasks, particularly among late-career researchers (66%), according to the ExplanAItions study.

Despite these challenges, researchers expect a rapid expansion in how AI is employed throughout the research process. Notably, many researchers already recognize AI’s potential with a majority stating that the technology outperforms humans in more than half of the 43 use cases evaluated in the study.

Concerns about AI models

While interest in AI remains high, concerns about the models persist. About 81% of researchers expressed one or more concerns about AI, with ethical issues (54%), lack of transparency in AI training and facility (46%), accuracy (51%), and data security or privacy (47%) ranking as the most common. This highlights the significant obstacles researchers face in increasing their use of AI, the report said.

Regional differences also emerged, with researchers in Japan (85%) and China (84%) reporting heightened concerns about the AI models. The study also found that social sciences researchers (86%) and researchers in the academic sector (83%) cited concerns regarding AI’s role and implications in their fields.

Where AI outperforms humans — and where humans shine

AI is recognized for excelling in tasks requiring speed, accuracy, and processing large datasets. In those instances, the researchers “recognize that AI can save time, handling monotonous and repetitive work,’’ the study found.

However, human expertise remains superior in areas requiring intuition, judgment, creativity, and complex problem-solving.

“We’ve heard researchers loud and clear,’’ said Jay Flynn, Wiley executive vice president and general manager of research and learning, in a statement. “We’re committed to supporting authors as they navigate this transformation and will offer guidance on how to use generative AI tools with greater confidence.”

The post AI Usage By Researchers Varies Depending on Their Career Phase, Field & Region appeared first on eWEEK.

The news sent the AI company’s stock surging nearly 35% in the past week and 211% over the past year, reflecting strong investor confidence in government contracts, according to Investing.com.

Despite the strength of BigBear.ai’s stock, InvestingPro analysis indicates the stock “currently appears overvalued relative to its Fair Value, and analysts do not expect profitability this year,’’ the company said. “BigBear.ai’s future performance may also be affected by competition and other risks outlined in the company’s SEC filings.”

AI system aims to flag and predict adversaries’ media

Under the terms of the contract, BigBear.ai will develop its Virtual Anticipation Network (VANE) prototype for the CDAO. Using AI and machine learning, the VANE system is designed to aggregate and analyze a wide variety of data points. The goal is to assist the CDAO’s ability to identify key trends and topics related to areas of interest by potential foreign adversaries. The system is also designed to predict adversarial activities in multi-domain environments and provide clear insights to support U.S. military and intelligence operations.

“By advancing VANE within CDAO, we are arming our warfighters with sophisticated intelligence capabilities to leverage foreign insights critical to the safety of our Nation and those protecting it,” said Ryan Legge, president of national security at BigBear.ai in a statement.

The VANE system received the “awardable” status on the CDAO Tradewinds Solutions Marketplace in April 2024.

The contract includes a transition plan for VANE with the goal of future deployment on the CDAO’s Advana enterprise data and analytics platform, which proves access to a large set of DoD data across various systems. This would expand access to the advanced AI capabilities across the DoD’s Combatant Commands.

VANE pegged for additional government services

BigBear.ai was also recently awarded a DoD contract to use the VANE prototype for geopolitical risk analysis, and the company has secured a contract under the Department of the Navy’s SeaPort Next Generation program to provide technical and engineering services to the U.S. Navy and other federal agencies, according to Investing.com

The post BigBear.ai Wins DoD Contract, Sending Its Stock Soaring appeared first on eWEEK.

As of January 2024, the San Francisco-based company had nearly 73,000 workers. Approximately 51% of Salesforce employees are located in the U.S., and 49% are located internationally.

Salesforce is expected to report its fourth-quarter earnings by the end of February.

Salesforce’s AI focus and open sales roles

The customer relationship management software company has not said where the internal cuts will take place; however, in tandem with the reported layoffs, the company is currently actively hiring salespeople to promote its AI products to customers.

Salesforce has signed more than 1,000 paid deals for its platform Agentforce, which creates AI-powered virtual representatives, according to CEO Marc Benioff. In a December 2024 podcast, Benioff said Salesforce would not be hiring more software engineers this year because of productivity gains of more than 30% from AI.

What is Agentforce?

Agentforce agents operate autonomously and retrieve data on demand to build action plans for any task and execute it without the need for human involvement. For example, an agent can use customer engagement data to identify an opportunity to upsell and generate a personalized email to a prospect. It is natively integrated into the entire Salesforce Customer 360 platform.

Agentforce “is the only thing that really matters today,’’ Benioff said in the podcast.

Companies using the platform include Wiley, OpenTable, Saks, and ezCater, according to Salesforce.

Other tech company layoffs

Massive layoffs began at tech companies in 2023, and the trend has continued. So far this year, Amazon, Microsoft, and Facebook parent company Meta have all moved to trim their corporate workforces. Microsoft targeted employees it classified as low performers, Business Insider reported.

In late 2024, layoffs occurred at various tech organizations including Yahoo, Calendly, AMD, iRobot, Mozilla, and Akamai.

Independent layoffs tracker Layoffs.fyi reported that so far this year, 31 tech companies have laid off just over 7,000 employees. There were more than 150,000 job cuts among 542 companies in 2024, and 264,000 in 2023, the site reported.

The post Salesforce Reportedly Laying Off 1K Workers While Adding AI-Focused Sales Roles appeared first on eWEEK.

How does OpenAI’s deep research tool work?

Once ChatGPT is given a prompt, OpenAI’s deep research agent will find, analyze, and pull together hundreds of online resources to create a comprehensive report that the company claims is on par with that of a research analyst.

“What you get from deep research is a comprehensive, fully cited research paper, essentially something that an analyst or an expert in a field might produce to you,” OpenAI Chief Research Officer Mark Chen said during a demo of the new deep research tool.

The agent uses reasoning to search, interpret, and analyze vast amounts of online text, images, and PDFs. Deep research will pivot when applicable to react to information it encounters, OpenAI said.

The agent was trained on real-world tasks requiring a browser and Python, “using the same reinforcement learning methods behind OpenAI o1, our first reasoning model,’’ the company said.

Although o1 is capable when it comes to coding, math, and other technical domains, OpenAI said deep research builds on reasoning capabilities to solve many real-world work and everyday challenges that demand more context and information gathering from diverse online sources.

The company issued a caveat that deep research is still in the early stage of development and “can sometimes hallucinate facts in responses or make incorrect inferences,” though at a lower rate than other ChatGPT models.

Who is the market for OpenAI’s deep research tool?

Deep research is aimed at people who do intensive work in areas including finance, science, policy, and engineering. However, shoppers who want very personalized recommendations on purchases that typically require a lot of research, such as cars, appliances, and furniture, will also find it useful, the company stated.

OpenAI’s Pro customers who pay $200 a month are the first people to be able to use this deep research tool.

The post OpenAI Launches Deep Research Tool For Complex Web Tasks – With A Caveat appeared first on eWEEK.

Strengthening AI Across the Hybrid Cloud

As part of this collaboration, Red Hat is enhancing the availability of software such as RHEL AI and Red Hat OpenShift AI in AWS Marketplace, including “bring your own subscription” (BYOS) and private offers. The vendor said it will support AI accelerators and GPUs from chip providers such as AMD, Intel, and NVIDIA. Red Hat said its goal is to help provide organizations with ready-made AI capabilities that can be scaled more easily with Red Hat OpenShift AI on AWS, either self-managed or through the services’ built-in capabilities.

“As AI becomes the next critical enterprise IT decision, we’re making optionality a reality in accelerated compute infrastructure, enabling customers to select the hardware accelerators that make the most sense for their unique hybrid cloud AI strategies and workloads,” Red Hat’s Senior Vice President of Partner Ecosystem Success Stefanie Chiras said in a statement.

As AI Becomes a Priority, Modernized Infrastructure is Needed

Organizations are navigating major migrations for VMs across complex IT environments even as they grapple with uncertainty and rising costs in the management of their virtual infrastructure. At the same time, many are increasingly prioritizing AI to stay competitive.

The vendor said Red Hat OpenShift Service on AWS aims to make migration of VMs and containerized workloads to the cloud easier, as well as help organizations with their application modernization initiatives. Additionally, Red Hat OpenShift will run as a self-managed offering on AWS EC2 bare metal instances.

Enterprises need to think beyond just managing clusters of virtual machines and consider how to optimize their virtualized infrastructure “as an engine for modernization that supports their artificial intelligence roadmaps, modern applications, and prepare for the next generation of IT, IDC Research Director Gary Chen said in a statement.

Explore our list of the top companies shaping AI to see which vendors are redefining the industry and the technology.

The post Red Hat’s Hybrid Cloud AI Strategies Expand Through AWS Collaboration appeared first on eWEEK.

“I made the Indian version of Mona Lisa using AI,’’ she announced in her post. “Give her a name.”

Social Media Reacts

Viewers suggested names that illustrate the Indian adaption of the Mona Lisa, such as “Shona Lisa,” “Mona Tai,” and “Lisa Ben.” Other comments imagined how the masterpiece could look in different regional styles. Some comments went beyond the playful.

“To be frank, this is beautiful—more womanly than the original,” one viewer wrote. Sorry art connoisseurs, this is a layman’s opinion.” Perhaps inspired by the response, Pandey also posted an AI-created Indian version of Queen Elizabeth II on X and asked viewers which they preferred.

Will AI Art Kill Creativity?

More than just a fun post, Pandey’s Indian Mona Lisa speaks to a larger debate about generative AI and its role in artmaking. AI is comprised of two elements: algorithms and data. Artists who create AI-generated art say it is a tool or extension of themselves that enables them to expand creative boundaries as they develop work. Many artists feel strongly that human involvement and influence are still essential to the creative process.

Hugh Leeman, an artist and lecturer who lectures at Johns Hopkins University, Duke University, and Colorado State University, has created a body of work, half from his own imagination and the other half with the help of AI.

Leeman believes AI tools can be used “to better our lives,” adding that he uses them “to create images based on people’s unique lived experiences told in stories embedded with their emotions and memories.” At the same time, however, he fears that “the coming generations will likely lose elements of agency of their creativity through the further merging of AI with humans in our social ecosystem as well as our physiology.”

Many artists are also concerned about AI being used for plagiarism. Their work is used to train algorithms, but it can also be copied using AI tools.

The post The AI-Generated Mona Lisa That Stole the Show: Rashi Pandey’s Indian Twist on a Renaissance Classic appeared first on eWEEK.