Home Blog Page 228

$100 Billion Investment in US Chipmaking

0

Taiwan Semiconductor Manufacturing Co. to Invest $100 Billion in US Chip Manufacturing

Largest Chipmaker to Expand Operations in Arizona

Taiwan Semiconductor Manufacturing Co. (TSMC), the world’s largest independent semiconductor foundry, has announced that it will invest at least $100 billion to expand its chip manufacturing operations in the United States. The company will build two additional chip manufacturing facilities in Phoenix, Arizona, as part of its efforts to strengthen its presence in the global market.

Large-Scale Investment

The $100 billion investment builds upon the $65 billion TSMC has already committed to building three Arizona factories, as well as the $6.6 billion the Biden administration awarded to TSMC under the CHIPS Act. The company has already begun producing 4-nanometer chips at its Arizona plant, which is expected to produce chips using "2nm or even more advanced process technology" by the end of the decade.

Timeline and Job Creation

Last year, TSMC pushed back the timeline for its second Arizona plant, saying it will open in 2027 or 2028 instead of 2026. The company’s CEO, C.C. Wei, emphasized that the expansion will create thousands of high-paying jobs and produce many AI chips. The success of TSMC’s first plant in Arizona, which produces the most advanced chip made on US soil, has been a significant factor in the company’s decision to expand its operations in the region.

TSMC’s Plans and Milestones

  • The company plans to produce chips using "2nm or even more advanced process technology" by the end of the decade.
  • The second Arizona plant is expected to open in 2027 or 2028.
  • The expansion will create thousands of high-paying jobs.
  • TSMC is producing the most advanced chip made on US soil at its first Arizona plant.

FAQs

Q: What is TSMC’s new investment in the US?
A: TSMC is investing at least $100 billion to expand its chip manufacturing operations in the US.

Q: Where will TSMC build its new facilities?
A: TSMC will build two additional chip manufacturing facilities in Phoenix, Arizona.

Q: What is the expected timeline for the new facilities?
A: The second Arizona plant is expected to open in 2027 or 2028.

Q: How many jobs will be created as a result of the expansion?
A: The expansion is expected to create thousands of high-paying jobs.

Q: Why is TSMC expanding its operations in the US?
A: TSMC is expanding its operations in the US to strengthen its presence in the global market and produce advanced chips using "2nm or even more advanced process technology" by the end of the decade.

Amazon Slammed by Judge in ‘Landmark’ Logo Dispute Ruling

0

Amazon Fined Record $39M in Trademark Dispute

Beverly Hills Polo Club vs. Amazon: A Tale of Logo Infringement

Amazon has been slapped with a record fine of $39 million (339 crore rupees) by an Indian court for infringing on the "Beverly Hills Polo Club" (BHPC) trademark. The online retail giant was sued by Lifestyle Equities, the owner of the BHPC brand, in 2020. The dispute revolves around an Amazon-affiliated seller called Cloudtail India, which sold apparel featuring a logo that is nearly identical to BHPC’s.

The Logo Dispute: A Tale of Unoriginality

The logo in question features an illustration of a polo player on a horse, swinging a polo mallet. This may not sound particularly unique, as it is reminiscent of the Ralph Lauren logo. However, the logo used by Cloudtail India, under the brand name ‘Symbol’, is almost indistinguishable from BHPC’s logo.

The Court’s Ruling: A Blow to Amazon’s Reputation

The Delhi High Court ruled against Amazon, citing "deliberate and wilful infringement" by the company. The court emphasized Amazon Seller Services’ role as an intermediary and held it accountable for the infringement. The ruling is seen as a landmark judgment in the context of online trademark infringement.

Amazon’s Blame Game Falls Flat

Amazon attempted to shift the blame onto Cloudtail India, but the court was not swayed. Justice Prathiba M Singh described Amazon’s actions as a "deliberate strategy of obfuscation, pretending to wear different hats – one as an intermediary, one as a retailer, and one as a brand owner – all in an attempt to shift responsibility and evade liability for trademark infringement."

Conclusion

This case serves as a stark reminder of the importance of protecting one’s intellectual property. Even the largest companies can fall victim to logo disputes, and it is crucial to be vigilant and take swift action to protect one’s brand identity.

FAQs

Q: What was the nature of the logo dispute between Amazon and Beverly Hills Polo Club?
A: The logo in question features an illustration of a polo player on a horse, swinging a polo mallet, and is almost indistinguishable from BHPC’s logo.

Q: Who was held responsible for the trademark infringement?
A: Amazon Seller Services, an intermediary, was held accountable for the infringement by the Delhi High Court.

Q: What was the amount of the record fine imposed on Amazon?
A: $39 million (339 crore rupees)

Q: What was the outcome of the court’s ruling?
A: The court ruled against Amazon, citing "deliberate and wilful infringement" and emphasized Amazon’s role as an intermediary in the dispute.

Data Shows Google AI Overviews Changing Faster Than Organic Search

0

AIO Overviews and Organic Search Results: New Research Insights

Recent research by Authoritas has shed new light on the evolution of AI Overviews (AIO) and their relationship with organic search results. The study found that AIO exhibits more volatility than organic search results, changing at a faster rate. This volatility doesn’t correlate with organic search volatility, suggesting that AIO is replacing or enhancing search results.

AIO is designed to summarize answers to complex queries by combining data from multiple sources, creating a precise long-form answer. Organic search results, on the other hand, provide topically relevant but not necessarily precise answers. This difference in approach explains why AIO and organic search results change independently.

The study’s findings raise important questions about the role of AIO in SEO strategies. Instead of ranking for information gain, it’s more effective to optimize content to be concise and precise, similar to optimizing for featured snippets.

A complex query is a query that demands a precise answer, which may not exist on a single website. An example of a complex query is asking "how is men’s fashion influenced by military style?" Organic search results may not provide a precise answer, but AIO can by combining information from multiple sources.

The line between complex queries and AIO is blurry, and Google’s AIO is constantly changing. Adding a word like "what" or "how" to a query can trigger an AIO. For instance, the query "how is men’s fashion influenced by military style?" generates an AIO answer that summarizes information from multiple websites.

Organic search results contain topically relevant results, but may not answer the question. In contrast, AIO delivers a precise answer by combining information from multiple sources.

How to Optimize for AIO?

Optimizing for AIO is similar to optimizing for featured snippets: create concise and precise content. Instead of ranking for information gain, focus on creating content that is precise and informative. This approach is more effective than trying to rank for follow-up questions.

Conclusion

The research by Authoritas provides valuable insights into the evolution of AIO and its relationship with organic search results. By understanding the differences between AIO and organic search results, marketers and SEO professionals can develop more effective strategies for leveraging AIO in their marketing efforts.

FAQs

Q: What is a complex query?
A: A complex query is a query that demands a precise answer, which may not exist on a single website.

Q: What is the difference between AIO and organic search results?
A: AIO provides precise answers by combining information from multiple sources, while organic search results provide topically relevant but not necessarily precise answers.

Q: How do I optimize for AIO?
A: Optimize for AIO by creating concise and precise content, similar to optimizing for featured snippets.

You can now talk to Google Gemini from your iPhone’s lock screen.

0

Google Gemini Users Can Now Access AI Chatbot from iPhone’s Lock Screen

New Update Allows Users to Call Up Gemini Live from Lock Screen

Google has released an update that allows users of its Gemini app to access its AI chatbot, Gemini Live, directly from their iPhone’s lock screen. This feature was first spotted by 9to5Google. With this update, users can now call up Gemini Live, a relatively real-time voice feature for Google’s AI chatbot, before they even unlock their phone by adding a Gemini widget to their lock screen.

Additional Lock Screen Widgets

The updated Gemini app also includes several other lock screen widgets, including one for taking pictures using the iPhone camera and uploading them to Gemini, one for setting reminders and calendar events, and another that lets users jump straight to a text chat with Gemini.

Comparing with Other AI Assistants

As Apple’s version of an AI-enabled Siri reportedly faces delays until 2027, competitors in the AI space are stepping in to supply iPhone users with AI assistants of their own. These features may give iPhone users a sense of what’s possible with LLMs and voice assistants, though Apple’s version of an LLM-powered Siri may be far more integrated with the iPhone’s other functions when it ultimately ships.

ChatGPT’s iOS App Also Offers Similar Feature

ChatGPT’s iOS app also lets users call up OpenAI’s near real-time voice feature, Advanced Voice Mode, from the lock screen.

Project Astra Update

Google also announced on Monday that later in March, it will allow Gemini users on Android to ask its AI chatbot questions about video and onscreen content, and get answers in real-time. These features were first unveiled as part of Project Astra, Google DeepMind’s multimodal AI project that is slowly making its way into the Gemini app. To start, these features will be available for subscribers to Google’s $20-a-month Gemini Advanced plan.

Conclusion

The latest update to the Gemini app brings a range of new features to iPhone users, including the ability to access Gemini Live from the lock screen. This update is part of Google’s efforts to compete with other AI assistants in the market, including Apple’s Siri and ChatGPT’s Advanced Voice Mode.

FAQs

  • What is Gemini Live?
    • Gemini Live is a relatively real-time voice feature for Google’s AI chatbot.
  • What other lock screen widgets are available in the updated app?
    • The app includes widgets for taking pictures, setting reminders and calendar events, and jumping straight to a text chat with Gemini.
  • Is the new feature available for Android users?
    • No, the new feature is currently only available for iPhone users. However, Google has announced that it will be available for Android users later in March.
  • What is Project Astra?
    • Project Astra is a multimodal AI project developed by Google DeepMind that is slowly making its way into the Gemini app.
  • How much does the Gemini Advanced plan cost?
    • The Gemini Advanced plan costs $20 per month.

Open AI and Anthropic Invite US Scientists to Experiment with Frontier Models

Partnerships between AI Companies and the US Government Expand, Raising Questions about AI Safety and Regulation

AI Jam Session Brings Scientists Together

On Friday, 1,000 scientists from nine labs gathered for an AI Jam Session hosted by OpenAI and Anthropic, testing their latest AI models. The event aimed to advance scientific research and assess the potential of AI to solve complex scientific challenges. The participating scientists were given access to several models, including OpenAI’s o3-mini and Claude 3.7 Sonnet, and Anthropic’s latest release, to evaluate their capabilities and provide feedback to improve future AI systems.

Government Partnerships

The AI Jam Session is part of existing agreements between the US government and AI companies. In April, Anthropic partnered with the Department of Energy (DOE) and the National Nuclear Security Administration (NNSA) to test Claude 3 Sonnet’s ability to reveal sensitive information. OpenAI also partnered with the DOE National Laboratories to "supercharge" their scientific research using its latest models.

The National Labs

The National Laboratories is a network of 17 scientific research and testing sites across the country, investigating topics such as nuclear security, climate change, and renewable energies. The partnership aims to accelerate and diversify disease treatment and prevention, improve cyber and nuclear security, and advance physics research.

Future of AI Safety and Regulation

The agreements between AI companies and the government raise concerns about the future of AI safety and regulation. The Trump administration’s AI Action Plan has yet to be announced, and staff cuts at the AI Safety Institute have been rumored. The head of the Institute has already stepped down, leaving the future of AI oversight in limbo. The risk of less oversight into how powerful and safe new models are, as deployment quickens, is a concern.

Conclusion

The partnerships between AI companies and the US government are expanding, but the future of AI safety and regulation remains unclear. As the development of AI continues to accelerate, it is crucial to prioritize safety and oversight to ensure responsible innovation.

Frequently Asked Questions

Q: What is the AI Jam Session?
A: The AI Jam Session is an event where scientists from nine labs test and evaluate the latest AI models.

Q: What are the goals of the AI Jam Session?
A: The goals are to advance scientific research, assess the potential of AI to solve complex scientific challenges, and provide feedback to improve future AI systems.

Q: What are the partnerships between AI companies and the US government?
A: The partnerships include OpenAI’s partnership with the DOE National Laboratories and Anthropic’s partnership with the DOE and NNSA.

Q: What is the future of AI safety and regulation?
A: The future of AI safety and regulation is uncertain, with the Trump administration’s AI Action Plan yet to be announced and staff cuts at the AI Safety Institute rumored.

How artificial intelligence redefines automation

0

Article contributed by automatica 2025

The adoption of artificial intelligence is an increasingly critical factor for the viability of industrial production.

European companies run a particular risk of getting left behind with regard to this cutting-edge technology – even though all kinds of industrial AI solutions are readily available and implementing them is now easier than ever. The leading exhibition automatica will impressively demonstrate this in June 2025.

A December 2024 survey by the Statista Research Department details just how precarious the situation is. For example, China has the highest prevalence of AI production technology at 94 percent.

The US comes in second, but is surprisingly far behind. 46 percent; that is, less than half of all manufacturing companies, use AI here. According to the survey, the German speaking countries (DACH region) come in last at only 20 percent.

Christian Fenk, CSO of the robominds AI company in Munich, feels that things should be different: “On the supplier side, Europe is among the global market leaders for AI solutions geared towards production.

“Anyone doubting this should attend automatica in Munich and see for themselves the wide range of available AI solutions covering all aspects of automation. Any company delaying their own adoption of this technology puts their competitive positioning at risk.”

US President Donald Trump is one of those who have understood the significance of AI. He started his second term by initiating the $500 billion Stargate venture intended to bring AI infrastructure in the US to a whole new level.

The enormous investment into this key technology is to not just ensure a competitive edge for the industry, but to secure the entire nation’s prosperity and sovereignty. What that means for Europe: Time to act.

Doing the impossible with smart automation

Let’s go back from global politics in Washington to a production hall in northern Germany: AI plays a major role here, too. It doesn’t affect the world order here, but it helps develops the competitive positioning of a plastics processing business facing the challenge of positioning a large variety of pre-separated components for assembly.

Conventional automation is not an option here as the enormous variety of components would require both a tremendous programming effort and continuous adjustments to the program as it is executed.

robominds was able to solve this problem using an AI solution that autonomously reacts to changes as they occur. A combination of robots, robobrain®, and suitable AI skills enables the system to recognize, grip, and separate all component variants without any programming or teaching effort.

“Real artificial intelligence acting as a link to cover unstructured processes unlocks entirely new fields of application in automation. Even if the customer’s product variety increases further, AI puts them in a comfortable spot as it enables flexible reactions to future changes,” says Tobias Rietzler, CEO at robominds.

Smart 3D vision replaces teaching and programming

The combination of 3D vision and powerful AI is one of the prerequisites for implementing smart robotics solutions. This technology enables robots to act appropriately in any given situation and to take on tasks subject to dynamic changes. This helps overcome the limitations of inflexible programmed sequences and lets machines achieve maximum autonomy.

At automatica, lots of machine vision providers present actual use cases for such solutions, including start-ups such as Mech-Mind Robotics. This company was founded in 2016 and, with the support of Intel and other investors, has attracted a total funding of more than $200 million.

It is already considered one of the top players in deploying AI and Deep Learning applications to implement extremely challenging automation tasks.

Trade fair stands of exhibitors such as Basler, Carl Zeiss, IDS, MVTec, or VMT offer visiting professionals the perfect opportunity to get up to speed on the latest technology in AI-based machine vision.

Our exhibitors will be happy to elaborate on tasks to be solved, the ease of system integration, associated costs, and return of investment.

Global robot manufacturers showcase visionary developments

Visitors are also eagerly awaiting the innovative solutions showcased by robot manufacturers at automatica. One thing we already know: We never had more registered robot manufacturers than this year, and many of our first-time exhibitors are from Asia.

Language programming is another AI-related topic that will play a pivotal role very soon. If it was possible to program robots using natural language, the greatest obstacle associated with their use would just vanish in an instant. A team from Augsburg, Germany, has shown that this dream of convenient language programming is now within reach.

The team at Kuka has been working on using generative AI to create program code for some time now. Roland Ritter, head of software portfolio management at Kuka, explains what this is all about: “We are currently developing an AI chatbot capable of translating natural language commands into code, which is then used to program the robot for the task at hand. If we succeed, anyone could perform entry-level robot programming with ease.”

Experiments are conducted in a virtual environment for now, and AI-generated robot programs are still being tested using digital twins in an effort to make them suitable for real-world applications. But progress is being made and it is only a matter of time before AI assistants and robot programming will work hand in hand.

Autonomous mobile robots – powered by artificial intelligence

Just as in stationary robots, artificial intelligence plays a key role in mobile robotics, too. Autonomous navigation may have the greatest impact in this domain as it unlocks fully autonomous deployments of AMRs in complex and continuously evolving environments.

It was not without good reason that ABB Robotics acquired Sevensense Robotics in 2023. The Swiss company specializes in VSLAM technology (Visual Simultaneous Localization and Mapping). This AI-based technology is considered a game changer as it enables AMRs to map unknown environments and navigate them with great precision.

Sami Atiya, head of robotics and discrete automation business area at ABB, says: “Each robot is equipped with machine vision technology and AI, and is tasked with scanning a specific part of the building. Each robot’s field of vision is used to compile a complete map so that AMRs can autonomously work in fast-changing environments.”

The numerous AGV and AMR providers use different navigation systems, and each of them can take on specific logistics tasks—learn all about it at automatica.

Both the choice of solutions and the market are huge. And that extends to the entire product range from grippers to cobots, and across all exhibition areas: AI is now ubiquitous and unlocks quantum leaps in terms of efficiency and economic potential.

Terraform for Beginners: Writing Your First Infrastructure Code

Introduction

Infrastructure as Code (IaC) has revolutionized cloud infrastructure management, allowing developers and DevOps engineers to define and manage infrastructure using code. Terraform, an open-source tool developed by HashiCorp, is one of the most widely used IaC tools. If you’re new to Terraform, this guide will help you write your first infrastructure code and deploy a simple AWS instance.

What is Terraform?

Terraform is an Infrastructure as Code (IaC) tool that enables you to define cloud and on-premises infrastructure in a declarative configuration file. It supports multiple providers such as AWS, Azure, Google Cloud, Kubernetes, and more.

Why Use Terraform?

  • Declarative Approach: Define the desired state, and Terraform manages the provisioning.
  • Multi-Cloud Support: Use the same tool for AWS, Azure, and other cloud providers.
  • State Management: Keeps track of resources via a state file.
  • Modularity: Reusable configurations make managing infrastructure easier.

Setting Up Terraform

Prerequisites

Before writing your first Terraform script, ensure you have:

  • An AWS account
  • Installed Terraform (Download here)
  • Installed AWS CLI and configured it with your AWS credentials

Writing Your First Terraform Code

Step 1: Create a Working Directory

Create a new directory for your Terraform project:

mkdir terraform-demo && cd terraform-demo

Step 2: Define the AWS Provider

Create a file named `main.tf` and add the following:

provider "aws" {
  region = "us-east-1"
}

Step 3: Define an EC2 Instance

Add a resource block to create an EC2 instance:

resource "aws_instance" "web" {
  ami           = "ami-0c55b159cbfafe1f0"  # Amazon Linux 2 AMI ID
  instance_type = "t2.micro"
}

3. Modularize Your Code

As your infrastructure grows, modularize your Terraform code to keep it organized and reusable. Create separate modules for different components like networking, compute, and storage.

4. Secure Your State File

The Terraform state file contains sensitive information. Always store it securely, preferably in a remote backend like AWS S3 with encryption enabled.

5. Use Terraform Workspaces

Terraform workspaces allow you to manage multiple environments (e.g., dev, staging, prod) within the same configuration:

terraform workspace new dev
terraform workspace new staging
terraform workspace new prod

6. Leverage Terraform Cloud

For team collaboration and advanced features like remote state management, policy enforcement, and cost estimation, consider using Terraform Cloud.

7. Continuous Integration/Continuous Deployment (CI/CD)

Integrate Terraform with CI/CD pipelines to automate the deployment and management of your infrastructure. Tools like Jenkins, GitLab CI, and GitHub Actions can be used for this purpose.

8. Stay Updated

Terraform and its providers are constantly evolving. Regularly update your Terraform version and provider plugins:

terraform init -upgrade

Common Pitfalls to Avoid

  • Hardcoding Sensitive Information: Use environment variables or secret management tools instead.
  • Ignoring State File Management: Always back up and manage it securely.
  • Overlooking Dependency Management: Use `depends_on` where necessary.
  • Not Using Remote Backends: Always use remote state storage for team projects.
  • Skipping `terraform plan` Before Applying Changes: This helps avoid unintended modifications.

Advanced Topics to Explore

  • Terraform Modules
  • Terraform State Manipulation
  • Policy as Code with Sentinel
  • Terraform Providers
  • Best Practices for Terraform

Conclusion

Terraform is a powerful tool that simplifies infrastructure management through code. By following this guide, you’ve taken the first step towards mastering Terraform. As you continue your journey, explore advanced features, best practices, and real-world use cases to become proficient in managing infrastructure with Terraform.

Call to Action

If you found this guide helpful, please share it with your network. For more tutorials and tips on DevOps, cloud computing, and infrastructure management, follow our blog and join our community. Let’s build and manage infrastructure better, together!

The Best Fitness Trackers and Watches

0

What We’re Looking For

Audience: Who is this fitness tracker for? The ideal fitness tracker for hardcore athletes will look different than the best one for casual users looking to get a few more steps in.

Battery life: A fitness tracker should be able to go at least two to three days between charges. If it’s a flagship smartwatch, it should at least offer quick charging.

Form factor: Is it a band or a smartwatch? Is it comfortable to wear 24/7?

Metrics: What metrics does this device track? We prioritize active minutes over steps and calorie burn, but health metrics like resting heart rate, VO2 Max, and sleep quality are pluses.

Consistency: Accuracy is nice, but it’s more important for measuring progress that your device delivers consistent results for heart rate, distance tracking, and steps.

Platform: Certain trackers are limited to specific phone ecosystems — others will work regardless of what your phone is. We prioritize the latter wherever possible.

Best Fitness Tracker Overall

The Whoop 4.0 is best for elite athletes or people who don’t mind experimental trackers. Photo by Victoria Song / The Verge

How We Test Fitness Trackers

Fitness trackers are meant to help you keep track of your health and activity. We do a mix of benchmark testing and experiential, real-life testing. That means snoozing with them, taking them out on GPS activities like runs and hikes, working up a sweat in several workouts, and comparing how they do against long-term control devices for heart rate, sleep, and GPS accuracy. Some factors we consider in our rankings are durability, performance, accuracy versus consistency in metrics, and of course, battery life.

Update, March 3: Updated pricing and availability.

Conclusion

In conclusion, the best fitness tracker for you will depend on your specific needs and preferences. By considering the factors mentioned above, you can find the perfect tracker for your lifestyle and fitness goals.

Frequently Asked Questions

Q: What is the best fitness tracker for beginners?
A: For beginners, we recommend the Fitbit Inspire series, which offers a simple and user-friendly interface.

Q: What is the best fitness tracker for athletes?
A: For athletes, we recommend the Garmin Forerunner series, which offers advanced features like GPS tracking and heart rate monitoring.

Q: How do I choose the right fitness tracker?
A: To choose the right fitness tracker, consider your fitness goals, budget, and personal preferences. Read reviews and do research to find the best tracker for you.

CoreWeave IPO redefines big tech’s “power law”

Unlock the Editor’s Digest for free

Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.

The Power Law: A Tale of Concentrated Returns

Venture capitalists have a concept they call the ‘power law’ – the idea that the bulk of returns in a portfolio comes from just a few investments. CoreWeave, a US data-center operator that has filed for an initial public offering, offers a new variation on this theme.

The CoreWeave Story

CoreWeave provides computing capacity for companies who want to train and use artificial intelligence. Simply put, it rents out Nvidia microchips, and lots of them. CoreWeave’s 250,000-plus processors in 32 data centers are more than double what Elon Musk’s supercomputer, Colossus, had at the end of 2024. Demand is brisk: by the end of last year, revenue was growing at an annualized 170%.

A Tale of Concentrated Returns

When it goes public, CoreWeave will reflect the classic power law in that its original investors will have done very well indeed. Assume its ebitda more than doubles to $3bn this year. Pop that on the same 15-times multiple that AI ‘hyperscalers’ Meta Platforms and Amazon enjoy, strip out around $6.5bn of net debt, and the company’s equity – not including new money raised – could be worth close to $40bn.

The Risks

The catch is that the notion of big-overshadows-small, familiar in VC portfolios, manifests at CoreWeave in less helpful ways too. There’s the outsized might of a handful of suppliers. Three of them – chiefly Nvidia, presumably – supplied three-quarters of CoreWeave’s material and product purchases in 2024. Its customer base is even more skewed: Microsoft alone accounted for 62% of revenue in 2024.

The Future

Close ties to Nvidia and Microsoft are, of course, the secret sauce. But any big dependency is a valuation risk. Microsoft procures data center space not just for itself but for ChatGPT owner OpenAI. But Sam Altman’s company is also building its own giant data centers. While CoreWeave sees a nearly-$400bn market by 2028, the eventual calculus of who needs what from whom is still up for grabs.

Conclusion

This isn’t a problem for now – CoreWeave reckons it has around $8bn of revenue due over the next two years, versus less than $2bn in 2024. But unless it diversifies fast, the company’s new investors will spend much time nervously watching for signs that the relationship remains intact.

Frequently Asked Questions

Q: What is CoreWeave?
A: CoreWeave is a US data-center operator that provides computing capacity for companies who want to train and use artificial intelligence.

Q: What is the current valuation of CoreWeave?
A: The company’s equity – not including new money raised – could be worth close to $40bn.

Q: What are the risks associated with investing in CoreWeave?
A: The company’s dependence on a few big customers and suppliers is a valuation risk, and its shareholder base is lopsided, with insiders holding over 80% of the votes.

Q: How does the company plan to diversify its revenue streams?
A: CoreWeave plans to expand its customer base and reduce its dependence on a few large customers.

Q: What is the expected market size for AI data centers by 2028?
A: CoreWeave sees a nearly-$400bn market by 2028.

Researchers surprised to find less-educated areas adopting AI writing tools faster

Corporate and Diplomatic Trends in AI Writing

All Sectors Show Similar Adoption Patterns

According to the researchers, all sectors they analyzed (consumer complaints, corporate communications, job postings) showed similar adoption patterns: sharp increases beginning three to four months after ChatGPT’s November 2022 launch, followed by stabilization in late 2023.

Predictors of AI Writing Usage

Organization age emerged as the strongest predictor of AI writing usage in the job posting analysis. Companies founded after 2015 showed adoption rates up to three times higher than firms established before 1980, reaching 10–15 percent AI-modified text in certain roles compared to below 5 percent for older organizations. Small companies with fewer employees also incorporated AI more readily than larger organizations.

AI Adoption by Sector

When examining corporate press releases by sector, science and technology companies integrated AI most extensively, with an adoption rate of 16.8 percent by late 2023. Business and financial news (14–15.6 percent) and people and culture topics (13.6–14.3 percent) showed slightly lower but still significant adoption.

International Trends

In the international arena, Latin American and Caribbean UN country teams showed the highest adoption among international organizations at approximately 20 percent, while African states, Asia-Pacific states, and Eastern European states demonstrated more moderate increases to 11–14 percent by 2024.

Implications and Limitations

Limitations of the Study

In the study, the researchers acknowledge limitations in their analysis due to a focus on English-language content. Also, as we mentioned earlier, they found they could not reliably detect human-edited AI-generated text or text generated by newer models instructed to imitate human writing styles. As a result, the researchers suggest their findings represent a lower bound of actual AI writing tool adoption.

Implications for Society

The researchers noted that the plateauing of AI writing adoption in 2024 might reflect either market saturation or increasingly sophisticated LLMs producing text that evades detection methods. They conclude we now live in a world where distinguishing between human and AI writing becomes progressively more difficult, with implications for communications across society.

“The growing reliance on AI-generated content may introduce challenges in communication,” the researchers write. “In sensitive categories, over-reliance on AI could result in messages that fail to address concerns or overall release less credible information externally. Over-reliance on AI could also introduce public mistrust in the authenticity of messages sent by firms.”

FAQs

Q: What was the main finding of the study?

A: The study found that all sectors showed similar adoption patterns of AI writing, with sharp increases beginning three to four months after ChatGPT’s launch, followed by stabilization in late 2023.

Q: What was the strongest predictor of AI writing usage?

A: Organization age was the strongest predictor, with companies founded after 2015 showing adoption rates up to three times higher than firms established before 1980.

Q: Which sectors showed the highest adoption of AI writing?

A: Science and technology companies showed the highest adoption rate, at 16.8 percent, followed by business and financial news and people and culture topics.

Q: What were the implications of the study’s findings?

A: The study suggested that the growing reliance on AI-generated content may introduce challenges in communication, including the potential for messages to fail to address concerns or release less credible information, and public mistrust in the authenticity of messages sent by firms.