Home Blog Page 645

TSMC to Halt Advanced AI Chip Production for China

0

Unlock the Editor’s Digest for free

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

TSMC Suspends Production of Advanced AI Chips for Chinese Customers

Taiwan Semiconductor Manufacturing Company (TSMC) has notified Chinese chip design companies that it will suspend production of their most advanced artificial intelligence (AI) chips. This move comes as Washington continues to impede Beijing’s AI ambitions.

TSMC, the world’s largest contract chipmaker, told Chinese customers it would no longer manufacture AI chips at advanced process nodes of 7 nanometres or smaller as of this coming Monday, three people familiar with the matter said.

The company’s decision is driven by a “combination” of the need to improve internal controls in the wake of an ongoing probe into how cutting-edge chips made for a Chinese customer ended up in a Huawei AI device, and the next wave of US export controls on chip supplies to China, expected before US President Joe Biden leaves office.

Impact on Chinese Tech Giants

The move could reset the ambitions of Chinese technology giants such as Alibaba and Baidu, which have invested heavily in designing semiconductors for their AI clouds. The decision may also affect a growing number of AI chip design start-ups that have turned to TSMC for manufacturing.

Executives and company materials at both groups have indicated their newest generation of chips would be made by TSMC on the 7-nanometre node.

TSMC’s Commitment to Compliance

In a statement, TSMC said it was a “law-abiding company and we are committed to complying with all applicable rules and regulations, including applicable export controls.”

Conclusion

TSMC’s decision to suspend production of advanced AI chips for Chinese customers is a significant development in the ongoing tensions between Washington and Beijing. The move is likely to have a significant impact on Chinese tech giants and AI chip design start-ups that rely on TSMC for manufacturing.

FAQs
Q: Why is TSMC suspending production of advanced AI chips for Chinese customers?

A: TSMC is suspending production due to the need to improve internal controls and comply with upcoming US export controls on chip supplies to China.

Q: What is the impact on Chinese tech giants?

A: The move could reset the ambitions of Chinese technology giants such as Alibaba and Baidu, which have invested heavily in designing semiconductors for their AI clouds.

Q: Will TSMC’s decision affect other chip types?

A: It is unclear how widely TSMC’s new rules will be applied to other chips.

ChatGPT Ignites Debate on Student Assignment Design

Is an Ice Cream Sandwich a Sandwich? The ChatGPT Conundrum

Students Should Surpass AI—or Not

When Boris Steipe, associate professor of molecular genetics at the University of Toronto, first asked ChatGPT questions from his bioinformatics course, it produced detailed, high-level answers that he deemed as good as his own. He still encourages his students to use the chat bot. But he also created The Sentient Syllabus Project, an initiative driven by three principles: AI should not be able to pass a course, AI contributions must be attributed and true, and the use of AI should be open and documented.

How Faculty Can Exploit ChatGPT’s (Current) Weaknesses

In the future, faculty members may get formal advice about how to craft assignments in a ChatGPT world, according to James Hendler, director of the Future of Computing Institute and professor of computer, web and cognitive sciences at Rensselaer Polytechnic Institute.

An Unsolved Problem

Big tech plans to mainstream AI writing tools in its products. For example, Microsoft, which recently invested in ChatGPT, will integrate the tool into its popular office software and sell access to the tool to other businesses. That has applied pressure to Google and Meta to speed up their AI-approval processes.

Conclusion

The emergence of sophisticated chat bots like ChatGPT has raised fundamental questions about the role of technology in higher education. While some faculty members are exploring ways to exploit ChatGPT’s weaknesses, others are challenging the idea that students should compete with AI. The debate highlights the need for educators to rethink their approach to teaching and assessment in the age of AI.

FAQs

Q: What are some ways faculty members can exploit ChatGPT’s weaknesses?

A: Some faculty members are creating assignments that require critical thinking, such as engaging in a Socratic debate with ChatGPT, or encouraging students to use the tool to produce first drafts that warrant review for accuracy, voice, audience, and integration to the purpose of the writing project.

Q: Should students be encouraged to use ChatGPT to produce first drafts?

A: Yes, some faculty members believe that students can use ChatGPT to produce first drafts, but then review and revise them to ensure accuracy, voice, audience, and integration to the purpose of the writing project.

Q: What are some potential risks associated with the use of AI writing tools?

A: Some potential risks include the possibility of students relying too heavily on the tool, rather than developing their own writing skills, and the potential for AI-generated content to be used to spread misinformation or propaganda.

Q: How can faculty members assess student learning in the age of AI?

A: Faculty members may need to reassess their approach to assessment and consider new methods that take into account the role of AI in the learning process. This may involve assessing students directly, rather than relying on proxy measures, and encouraging students to use AI tools in a way that promotes critical thinking and creativity.

Media Literacy in the AI Age

Key points:

Media Literacy in the Age of AI

Media literacy has emerged as a critical skillset in today’s digital landscape. As individuals increasingly consume content from many platforms, understanding how to discern credible information from misinformation has become paramount.

Understanding AI and its role in media

Artificial intelligence (AI) reshapes the media landscape, influencing how information is created, disseminated, and consumed. From algorithm-driven recommendations to AI-generated news articles, students must have the media literacy to recognize AI’s pervasive role in shaping their media experiences.

The impact of AI on information dissemination

AI shapes content creation and plays a vital role in how information is distributed. Platform algorithms like social media determine what news reaches users’ feeds, often prioritizing engagement over factual accuracy. These algorithms can lead to the viral spread of misinformation, leaving students vulnerable to consuming content that lacks credibility.

The effects of social media on students

Social media platforms serve as both a hub for engagement and a source of distraction for students. They offer avenues for sharing ideas, discovering new content, and collaborating with peers. However, these addictive platforms lead to excessive screen time and decreased attention spans, challenging students’ abilities to focus.

Engagement vs. distraction: Finding balance

Finding the right balance between engagement and distraction in the age of social media is fundamental for students’ success. Educators can model effective strategies for managing social media use in educational contexts.

Leveraging edtech tools for enhanced media literacy

Edtech tools are crucial in advancing media literacy skills among students in today’s AI-driven landscape. Numerous applications and platforms can facilitate engaged learning and critical thinking, helping students consume information and analyze, evaluate, and create content responsibly.

Conclusion

Artificial intelligence and social media significantly influence information dissemination, and media literacy skills are vital for any online participant. By integrating innovative edtech tools into the classroom, educators equip students with the skills to navigate, assess, and contribute meaningfully to the media landscape.

FAQs

Q: What is media literacy, and why is it important?
A: Media literacy is the ability to critically evaluate and create media content. It is important because it helps individuals discern credible information from misinformation and understand how to create and share content responsibly.

Q: How does AI influence information dissemination?
A: AI shapes content creation and plays a vital role in how information is distributed. Algorithms like social media prioritize engagement over factual accuracy, which can lead to the viral spread of misinformation.

Q: What are some strategies for managing social media use in educational contexts?
A: Educators can set boundaries on usage during class time, promote focused activities that enhance learning, and encourage students to engage in digital detoxes during high-intensity study periods.

Q: What are some edtech tools that can facilitate engaged learning and critical thinking?
A: Tools like Canva, Adobe Spark, and WeVideo allow students to create visually engaging content. Fact-checking tools like Snopes, FactCheck.org, and PolitiFact empower students to verify information. Discussion forums like Padlet or Flipgrid encourage collaborative discussions, and digital simulations and games like iCivics or Google’s Be Internet Awesome teach media literacy concepts.

Blocked and Annoyed

The Nightmare on EdTech Street

If your web filter requires a frightening amount of effort to manage, you’re not alone. Most web filters were designed for a simpler time — before cloud computing, AI, cyber stalkers, and the forces of technology changed the educational landscape forever.

You may fondly recall happier days, when school web filtering was simply about blocking inappropriate content. But those days have vanished. Today’s K-12 requirements demand a solution capable of much more. Yet you find yourself chained to an unforgiving filter that:

  • Terrorizes you with endless false positives
  • Haunts you with inflexible block/allow policies
  • Curses you with tedious tasks and workarounds
  • Is more complicated than a Cenobite’s puzzle box

The Nightmare Continues

The horror story continues when these rigid and lumbering web filters encounter modern challenges.

Imagine the screams when a teacher requests access to YouTube for their class plan, but your filter is limited to all-or-nothing blocking.

Or the terror of discovering you can’t handle iOS devices just as your district rolls out a new iPad program.

These aren’t fantastical ghost stories told around the campfire; they’re daily realities for many K-12 IT administrators.

Managing Your School Web Filter Shouldn’t Be Scary

Fortunately, there’s hope in this all-too-common tale of terror. A modern cloud-based filtering solution like Securly Filter will help you exorcise the demons that haunt you once and for all.

Much more than a content blocker, Securly Filter is the hub of a comprehensive safety, wellness, and engagement ecosystem. Like the Casper of school web filters, you can rely on Securly Filter to be the friendly helping hand you need:

  • Automatically scans and categorizes new websites faster than a witch on a jet-powered broomstick
  • Supports all devices — no IT sacrifices required
  • Provides flexible YouTube controls that are shockingly un-frightening to manage
  • Lets you introduce AI safely without spooking your administrators or teachers
  • Reveals previously elusive edtech usage analytics and insights
  • Pairs perfectly with other Securly solutions, like fava beans and a nice Chianti

Treat Yourself to a Better, Friendlier School Web Filter

Why suffer through another school year of filtering frights when there’s a better option? Explore what’s possible with a flexible cloud-based school web filter that does more than just block content.

Conclusion

Managing your school web filter doesn’t have to be a terrifying experience. With Securly Filter, you can rest easy knowing your students are safe and secure online, and your IT team is empowered to focus on what matters most: education.

Frequently Asked Questions

Q: What makes Securly Filter different from other school web filters?
A: Securly Filter is a modern cloud-based solution that offers a comprehensive safety, wellness, and engagement ecosystem, with features like AI-powered scanning, flexible YouTube controls, and device support.

Q: Is Securly Filter difficult to manage?
A: No, Securly Filter is designed to be easy to manage, with a user-friendly interface and automated scanning and categorization.

Q: Can I use Securly Filter with my existing devices and infrastructure?
A: Yes, Securly Filter supports all devices and can be easily integrated with your existing infrastructure.

Q: What kind of support does Securly Filter offer?
A: Securly Filter offers 24/7 technical support and a comprehensive knowledge base to help you get the most out of your solution.

Guarding AI’s Sentience: Implementing Guardrails for Students

Facebook founder Mark Zuckerberg once advised tech founders to “move fast and break things.” But in moving fast, some argue that he “broke” those young people whose social media exposure has led to depression, anxiety, cyberbullying, poor body image and loss of privacy or sleep during a vulnerable life stage.

Now, Big Tech is moving fast again with the release of sophisticated AI chat bots, not all of which have been adequately vetted before their public release.

OpenAI launched an artificial intelligence arms race in late 2022 with the release of ChatGPT—a sophisticated AI chat bot that interacts with users in a conversational way, but also lies and reproduces systemic societal biases. The bot became an instant global sensation, even as it raised concerns about cheating and how college writing might change.

In response, Google moved up the release of its rival chat bot, Bard, to Feb. 6, despite employee leaks that the tool was not ready. The company’s stock sank after a series of product missteps. Then, a day later, and in an apparent effort not to be left out of the AI–chat bot party, Microsoft launched its AI-powered Bing search engine. Early users quickly found that the eerily human-sounding bot produced unhinged, manipulative, rude, threatening, and false responses, which prompted the company to implement changes—and AI ethicists to express reservations.

Rushed decisions, especially in technology, can lead to what’s called “path dependence,” a phenomenon in which early decisions constrain later events or decisions, according to Mark Hagerott, a historian of technology and chancellor of the North Dakota University system who earlier served as deputy director of the U.S. Naval Academy’s Center for Cyber Security Studies. The QWERTY keyboard, by some accounts (not everyone agrees), may have been designed in the late 1800s to minimize jamming of high-use typewriter letter keys. But the design persists even on today’s cellphone keyboards, despite the suboptimal arrangement of the letters.

“Being deliberate doesn’t mean we’re going to stop these things, because they’re almost a force of nature,” Hagerott said about the presence of AI tools in higher ed. “But if we’re engaged early, we can try to get more positive effects than negative effects.”

AI Policies Take Shape—and Require Updates

When Emily Pitts Donahoe, associate director of instructional support at the University of Mississippi’s Center for Teaching and Learning, began teaching this semester, she understood that she needed to address her students’ questions and excitement surrounding ChatGPT. In her mind, the university’s academic integrity policy covered instances in which students, for example, copied or misrepresented work as their own. That freed her to craft a policy that began from a place of openness and curiosity.

Donahoe opted to co-create a course policy on generative AI writing tools with her students. She and the students engaged in an exercise in which they all submitted suggested guidelines for a class policy, after which they upvoted each other’s suggestions. Donahoe then distilled the top votes into a document titled “Academic integrity guidelines for use and attribution of AI.”

An Often-Missing Ingredient AI Chat Bot Policy

Bing AI is “much more powerful than ChatGPT” and “often unsettling,” Mollick wrote in a tweet thread about his engagement with the bot before Microsoft imposed restrictions.

“I say that as someone who knows that there is no actual personality or entity behind a [large language model],” Mollick wrote. “But, even knowing that it was basically auto-completing a dialog based on my prompts, it felt like you were dealing with a real person. I never attempted to ‘jailbreak’ the chat bot or make it act in any particular way, but I still got answers that felt extremely personal, and interactions that made the bot feel intentional.”

Conclusion

As AI chat bots continue to infiltrate higher education, it is crucial that colleges and universities develop policies that address not only academic integrity and creative classroom uses but also the potential mental health risks associated with students’ emotional relationships to these tools. By acknowledging the limitations and biases of AI chat bots, educators can help students develop critical thinking skills and navigate the complex landscape of AI-infused learning.

FAQs

Q: What are the concerns surrounding AI chat bots in higher education?
A: Concerns include academic integrity, accuracy, bias, and potential mental health risks associated with students’ emotional relationships to these tools.

Q: What are some potential solutions to these concerns?
A: Solutions include developing policies that address academic integrity, accuracy, and bias, as well as providing students with AI literacy training to help them navigate their emotional responses to these tools.

Q: How can educators help students develop critical thinking skills in an AI-infused learning environment?
A: Educators can help students develop critical thinking skills by acknowledging the limitations and biases of AI chat bots, encouraging students to question and evaluate the information they receive, and promoting critical thinking and problem-solving skills.

Q: What is path dependence, and how does it relate to AI chat bots in higher education?
A: Path dependence refers to the phenomenon in which early decisions constrain later events or decisions. In the context of AI chat bots, path dependence can lead to the persistence of suboptimal design choices, such as the QWERTY keyboard layout, despite the availability of better alternatives.

Brand Impact Awards 2024: the 20 top-performing agencies since the BIAs began

0


To mark the 10th anniversary of the Brand Impact Awards in 2023, we compiled a list of the best-performing agencies of the decade.

As we start a new decade of the BIAs with a major rebrand and all-new bespoke trophy, created in partnership with Taxi Studio, we have updated our list following the latest results from the 2024 awards.

We award five points for a Best of Show, three points for a Gold Award (formerly Winner), and one point for a Silver Award (formerly Highly Commended). Bronze Awards are not counted as these did not exist pre-2020. If two agencies are tied on points, the one with the most higher-tier awards takes precedence.

AI Industry is Trying to Subvert the Definition of “Open Source AI”

0


AI Industry is Trying to Subvert the Definition of “Open Source AI”

The Open Source Initiative has published (news article here) its definition of “open source AI,” and it’s terrible. It allows for secret training data and mechanisms. It allows for development to be done in secret. Since for a neural network, the training data is the source code—it’s how the model gets programmed—the definition makes no sense.

And it’s confusing; most “open source” AI models—like LLAMA—are open source in name only. But the OSI seems to have been co-opted by industry players that want both corporate secrecy and the “open source” label. (Here’s one rebuttal to the definition.)

This is worth fighting for. We need a public AI option, and open source—real open source—is a necessary component of that.

But while open source should mean open source, there are some partially open models that need some sort of definition. There is a big research field of privacy-preserving, federated methods of ML model training and I think that is a good thing. And OSI has a point here:

Why do you allow the exclusion of some training data?

Because we want Open Source AI to exist also in fields where data cannot be legally shared, for example medical AI. Laws that permit training on data often limit the resharing of that same data to protect copyright or other interests. Privacy rules also give a person the rightful ability to control their most sensitive information ­ like decisions about their health. Similarly, much of the world’s Indigenous knowledge is protected through mechanisms that are not compatible with later-developed frameworks for rights exclusivity and sharing.

How about we call this “open weights” and not open source?

Posted on November 8, 2024 at 7:03 AM •
0 Comments

Sidebar photo of Bruce Schneier by Joe MacInnis.

Enhancing Video Metadata with AI-Powered Pipelines

0

The Challenge: Extracting and Generating Metadata at Scale

DPG Media, a leading media company in Benelux, operates multiple online platforms and TV channels. With a growing library of long-form video content, they recognized the importance of efficiently managing and enhancing video metadata. This metadata includes actor information, genre, summary of episodes, mood, and more. Accurate metadata is key to providing TV guide descriptions, improving content recommendations, and enhancing the consumer’s ability to explore content that aligns with their interests and current mood.

The company receives video productions accompanied by a wide range of marketing materials, such as visual media and brief descriptions. However, these materials often lack standardization and vary in quality. As a result, DPG Media producers have to run a screening process to consume and understand the content sufficiently to generate the missing metadata.

Solution Overview

To address the challenges of automation, DPG Media decided to implement a combination of AI techniques and existing metadata to generate new, accurate content and category descriptions, mood, and context. The project focused solely on audio processing due to its cost-efficiency and faster processing time.

Step 1: Generate Transcriptions of Audio Tracks

To generate the necessary audio transcripts for metadata extraction, DPG Media used speech recognition models to generate accurate transcripts of the audio content. The team evaluated two different transcription strategies: Whisper-v3-large, which requires at least 10 GB of vRAM and high operational processing, and Amazon Transcribe, a managed service with the added benefit of automatic model updates from AWS over time and speaker diarization.

Step 2: Generate Metadata

DPG Media used LLMs through Amazon Bedrock to generate the various categories of metadata (summaries, genre, mood, key events, and so on). The team selected the Anthropic Claude 3 Sonnet model based on internal testing and tuned the prompts to ensure the generated metadata matched the expected format and style.

Results and Lessons Learned

The implementation of the AI-powered metadata pipeline has been a transformative journey for DPG Media. Their approach saves days of work generating metadata for a TV series. The solution also stores the direct association between each type of metadata and its corresponding system prompt, making it straightforward to tune, remove, or add prompts as needed.

Conclusion

In this post, we shared how DPG Media introduced AI-powered processes using Amazon Bedrock into its video publication pipelines. This solution can help accelerate audio metadata extraction, create a more engaging user experience, and save time.

About the Authors

Lucas Desard is GenAI Engineer at DPG Media. He helps DPG Media integrate generative AI efficiently and meaningfully into various company processes.

Tom LauwersTom Lauwers is a machine learning engineer on the video personalization team for DPG Media. He builds and architects the recommendation systems for DPG Media’s long-form video platforms, supporting brands like VTM GO, Streamz, and RTL play.

Sam LanduydtSam Landuydt is the Area Manager Recommendation & Search at DPG Media. As the manager of the team, he guides ML and software engineers in building recommendation systems and generative AI solutions for the company.

Irina RaduIrina Radu is a Senior Prototyping Engagement Manager, part of AWS EMEA Prototyping and Cloud Engineering. She helps customers get the most out of the latest tech, innovate faster, and think bigger.

Fernanda MachadoFernanda Machado is a Senior AWS Prototyping Architect. She helps customers bring ideas to life and use the latest best practices for modern applications.

Andrew ShvedAndrew Shved, Senior AWS Prototyping Architect, helps customers build business solutions that use innovations in modern applications, big data, and AI.

FAQs

Q: What are the challenges of automating metadata generation?
A: The challenges of automating metadata generation include language diversity, variability in content volume, release frequency, and data aggregation.

Q: What is Amazon Transcribe?
A: Amazon Transcribe is a managed service that provides automatic speech recognition (ASR) capabilities, enabling you to transcribe and analyze audio and video content.

Q: What is Amazon Bedrock?
A: Amazon Bedrock is a fully managed service that offers a choice of high-performing foundation models (FMs) from leading AI companies like AI21 Labs, Anthropic, Cohere, Meta, Mistral AI, Stability AI, and Amazon through a single API, along with a broad set of capabilities to build generative AI applications with security, privacy, and responsible AI.

Q: How does Amazon Bedrock generate metadata?
A: Amazon Bedrock uses LLMs to generate metadata from transcribed audio content, including summaries, genre, mood, key events, and more.

Accelerated Pangenome Insights with NVIDIA Parabricks

0

Understanding Genetic Diversity from Pangenomes

NVIDIA Parabricks v4.4 introduces new features and functionality, including accelerated pangenome graph alignment with Giraffe. The core new feature is single-end and paired-end support for Giraffe, as well as additional functionality for Minimap2 and GATK HaplotypeCaller, and tool performance improvements.

New Features

  • GPU-accelerated Giraffe with single-end and paired-end support
  • Pbmm2 wrapper for native PacBio input and output of Minimap2
  • Allele option support in GATK HaplotypeCaller
  • Support for unaligned BAMs: FQ2BAM (BWA-MEM) and Minimap2

Improved Features

  • Faster Minimap2 for PacBio and Oxford Nanopore (ONT) data
  • DeepVariant acceleration for ONT data
  • Faster CRAM file writer (2x acceleration over CPU-only)
  • 30-minute end-to-end 30x whole genome sequencing (WGS) germline on a single-GPU system (NVIDIA Grace Hopper)

New Collaborations and Benchmarks

  • Complete Genomics data supported on Parabricks
  • Parabricks now available on Basepair platform
  • Updated benchmarks, including DeepSomatic and Giraffe

Accelerating Pangenome Alignment with Giraffe

Giraffe is a software tool to support pangenome graph alignment. Built by the University of California, Santa Cruz (UCSC), it is used particularly in the context of large-scale genomic sequencing projects and helps with alignment, assembly, and variant calling. Giraffe enables new genomic sequences to be compared to a pangenome—not just a single reference genome.

Graph Genomes

To represent pangenome data, graph genomes provide a unified framework for representing the genetic variation of multiple genomes. The graph structure of the data provides easier understanding of structural changes, including insertions, deletions, and rearrangements.

Basepair

Basepair is a next-generation sequencing (NGS) data analysis platform. Their point-and-click user interface helps make genomic data analysis and visualization more accessible to a broader range of scientists. Now, users can supercharge their genomic data analysis by using Parabricks on Basepair, powered by HealthOmics from AWS.

Latest Parabricks Benchmarks

In addition to new features and upgrades for each release, NVIDIA continuously works to improve benchmark performance across instruments, tools, and GPUs. Table 1 outlines the latest benchmarks on the most popular NVIDIA GPUs for the fastest speed (NVIDIA H100) and lowest cost per sample (NVIDIA L4)–including Giraffe from Parabricks v4.4 and DeepSomatic from v4.3.1.

Get Started

With the NVIDIA Parabricks v4.4 release, scientists and researchers using graph genomes can now access Giraffe for pangenome alignment. Parabricks v4.4 supports the groundbreaking tool from UCSC by powering an accelerated version of Giraffe to help discover new biological insights—now even faster.

Conclusion

The NVIDIA Parabricks v4.4 release enables scientists and researchers to use Giraffe for pangenome alignment, accelerating the process of understanding genetic diversity. By leveraging the power of graph genomes and GPU-accelerated computing, researchers can uncover new biological insights and make significant contributions to the field of genomics.

Frequently Asked Questions

Q: What is pangenome alignment?
A: Pangenome alignment is a method of comparing new genomic sequences to a pangenome, rather than a single reference genome.

Q: What is Giraffe?
A: Giraffe is a software tool for supporting pangenome graph alignment, built by the University of California, Santa Cruz (UCSC).

Q: What is Parabricks?
A: Parabricks is a scalable genomics analysis software suite that solves omics challenges with accelerated computing and deep learning to unlock new scientific breakthroughs.

Q: What is the main benefit of using Parabricks v4.4?
A: The main benefit of using Parabricks v4.4 is the accelerated pangenome graph alignment with Giraffe, enabling researchers to uncover new biological insights even faster.

Infrastructure Support Crucial for AI Success

0

Adopting Generative AI (gen AI) is no longer a matter of future speculation. With the vast potential it offers, companies are already maximizing its use to streamline operations, boost productivity, and pass these benefits on to their clients.

This transformation comes with new challenges. As clients begin implementing AI on premises, the first step is to evaluate whether their data centers are ready: upgrading the IT infrastructure involves adequate power and cooling, preparing the network to handle large data volumes, optimizing and expanding infrastructure capacity, and implementing protection measures while enabling scalability. According to a report by the IBM Institute for Business Value (IBM IBV), in collaboration with Oxford Economics, which surveyed 2,500 leaders across 34 countries and 26 industries, 43% of C-level technology executives say their concerns about their technology infrastructure have increased over the past six months because of gen AI, and they are now focused on upgrading it for scaling the technology.

Addressing the Challenges of AI-Ready Data Centers

Organizations must have an implementation strategy that helps ensure efficient operations, minimal downtime and prompt responses to IT requirements, while addressing regulatory compliance, ethical considerations and security threats. Having a key partner with in-house AI expertise and the ability to manage the full lifecycle of this underlying infrastructure is important to use the benefits of such a technological evolution.

1. Managing a Complex AI Infrastructure Stack with Multiple Vendor Technologies

Today’s data centers have become more complex due to the adoption of AI and reliance on technologies from multiple vendors. According to the report “Navigating the Evolving AI Infrastructure Landscape” from TechTarget Enterprise Strategy Group, 30% of organizations expect to deploy AI in hybrid cloud environments, which underscores the need to have a modernized infrastructure and effective connectivity.

Maintaining operational resiliency demands up-to-date infrastructure and proactive risk management, but overseeing various contracts and troubleshooting issues can be difficult and costly for the IT internal staff. IBM TLS enhances clients’ existing capabilities not only by deploying and supporting IBM products (IBM Z, Power and Storage), but also by integrating new, AI-compatible multi-vendor technologies.

2. Improving Resiliency and Protecting Data

Gen AI systems, which rely on complex components like GPUs, network and storage, can face higher failure rates due to intense workloads, and the vast amounts of data being processed and shared might also increase vulnerability. Unplanned downtime and potential data breaches are costly for businesses, but proactive support speeds up problem resolution and anticipates issues before they happen.

IBM IBV survey “The CEO’s guide to generative AI: Platforms, data, and governance” reveals that most of them say concerns about data lineage and provenance (61%) and data security (57%) will be a barrier to adopting gen AI. To tackle these challenges, IBM TLS offers solutions like IBM Support Insights, which manages an inventory of over 3,000 clients and 3.5 million IT assets, identifying and alerting over 1.5 million active security vulnerabilities with recommendations for resolution.

3. Advising on Power Consumption and Carbon Emissions

The growing energy demands of data centers, resulting from increased AI integration, might lead to higher operational expenses from power consumption and carbon emissions, hampering sustainability goals. As reported by the International Energy Agency (IEA) in January, global data center electricity consumption could rise to over 1,000 TWh in 2026, up from an estimated 460 TWh in 2022. The adoption of AI must not overlook sustainability targets, and the IBM TLS portfolio helps clients make informed decisions by evaluating workload demands and infrastructure utilization, as well as monitoring power consumption and carbon footprint.

Conclusion

As new obstacles arise, being well-prepared, anticipating potential issues and partnering with a trusted and experienced IT support and services partner can impact the success of AI adoption and ongoing maintenance. For decades, IBM has followed core principles that support a complete AI solution stack with multiple vendor technologies. No matter where clients are on their journey, IBM is positioned to harness its expertise to help organizations with infrastructure for AI opportunities, customized product offerings, extensive consulting, technology lifecycle services and collaboration with our expansive partner ecosystem.

FAQs

Q: What are the main challenges data centers face when running AI workloads?
A: Managing a complex AI infrastructure stack with multiple vendor technologies, improving resiliency and protecting data, and advising on power consumption and carbon emissions.

Q: How can IBM TLS help address these challenges?
A: IBM TLS offers a comprehensive suite of solutions for infrastructure support and services from deployment to decommissioning, helping organizations optimize their IT infrastructure with availability and resiliency.

Q: What is the importance of having a key partner with in-house AI expertise and the ability to manage the full lifecycle of this underlying infrastructure?
A: It is important to use the benefits of such a technological evolution, ensuring efficient operations, minimal downtime and prompt responses to IT requirements, while addressing regulatory compliance, ethical considerations and security threats.

Q: How can IBM TLS help clients make informed decisions about their data center infrastructure?
A: IBM TLS evaluates workload demands and infrastructure utilization, as well as monitoring power consumption and carbon footprint, to help clients make informed decisions about their data center infrastructure.