AI Every time you pop an idle thought into ChatGPT or another AI interface, it triggers a far-flung journey. Your query—“What is this rash?” perhaps—races to a hulking data center in a place like Abilene, Texas, where it’s routed to the right stack of servers. The traffic directors of all this activity are a series of high-bandwidth network switches—devices that coordinate data between servers—made by companies like Broadcom. The AI wave has reshaped Broadcom’s semiconductor business; its switches are now designed to support the high-capacity needs of massive AI data centers, and its custom AI accelerator line is a fast-growing source of revenue. “Over the last three years, things have changed pretty dramatically with AI,” Peter Del Vecchio, product line manager for Broadcom’s Tomahawk family of data center switches, told us. “For example, we announced [Tomahawk 5] in August of 2022, and the main market at that time was just general data center networking. ChatGPT was announced three months later, and pretty much everything changed.” Fast-forward three years: Broadcom has inked a deal with OpenAI to jointly build 10 gigawatts’ worth of AI accelerators over the coming years, tapping both Broadcom’s custom AI accelerators and its networking infrastructure. It’s one of a flurry of infrastructure partnerships that OpenAI has forged over the past several months. On the heels of this new deal, Tech Brew took a behind-the-scenes tour of Broadcom’s network switch R&D labs, where engineers undertake the multiyear process of assembling these devices. Keep reading here.—PK | | |
|
|
Presented By Hyland It’s official, we’re coining a new acronym: COMO, aka cost of missing out. It’s something tech leaders should be thinking long and hard about when it comes to business data. In a recent interview on The Ravit Show, Hyland CEO Jitesh S. Ghai sounded the alarm on enterprise data, noting that 80%–85% is unstructured, untapped, and unmanaged. “The final frontier in enterprise data is the unstructured data and enterprise content. If you’re not unlocking it, you’re leaving an ocean of insight and potential revenue untapped,” Ghai said. Ghai went on to explain how Hyland’s Content Innovation Cloud™ turns that unstructured data into actionable intelligence. That, in turn, can give businesses a critical advantage and help leaders avoid missed opportunities. Check out the interview to learn what’s possible with content intelligence, the costs of missing out on it, and how it can change just about…everything. |
|
GREEN TECH It’s not a New York City heat wave if the utility doesn’t send customers alerts asking them to limit usage of large appliances and air conditioners. And New Yorkers aren’t alone: In response to mounting energy demand, utility companies across the country are asking customers to curtail their energy use. These alerts—and the ensuing energy management—are part of demand-response programs, or coordinated ways energy customers can help reduce stress on the grid by using less energy during specific intervals. Customers might be financially incentivized to participate, like by paying cheaper non-peak rates. Two recent reports on demand-response programs, run by energy management companies Renew Home and Parity, showed how homes and hotels can be assets to a swamped grid. Collective action at home: Both demand-response scenarios used software to curtail small amounts of individual energy use. In the Renew Home study, conducted this summer, 5 million homeowners with Google Nest thermostats opted in to an Energy Shift program in which home temperatures were raised between .5 degrees and 2 degrees Fahrenheit within a two-hour period over numerous randomized trials. Keep reading here.—TC | | |
|
|
Together With HSBC Discover founder-first banking for the future. HSBC Innovation Banking US is committed to helping tech leaders get started on the future faster by providing access to an extensive global network, dedicated sector experts, and a robust suite of banking products and services. Learn more. |
|
AI Deloitte Australia recently took an embarrassing blow to its reputation when a report it compiled for the Australian government turned out to contain numerous AI-generated hallucinations. The firm will have to refund part of the $290,000 the Aussie government paid for the report. Though the Deloitte Australia incident’s unusual in that it involves a Big Four firm, it may point to a broader issue with poorly edited or cut-and-pasted AI-generated content in the workplace. The phenomenon’s even got a catchy name: workslop. Workslop’s toll: A Stanford University and BetterUp Labs survey published in the Harvard Business Review found that 40% of 1,150 US employees across a range of industries said they’d received workslop from colleagues in the past month. Staff estimated that around 15% of the work product they get is workslop, and that they spent an average of nearly two hours dealing with each incident. Workslop “shifts the burden of the work downstream” to the people who receive it and who then must fact-check, correct, rewrite, or interpret it, the authors wrote. Staff lose trust in coworkers who send them work products with obvious errors or that lack context, the Stanford study found. Workslop can have “pretty detrimental” effects on culture, Asha Palmer, SVP of compliance solutions at Skillsoft, told CFO Brew. Workslop can ultimately lead to “reputational damage” for both the organizations and individuals involved, Palmer said. Keep reading here.—CV | | |
|
|
Together With Atlassian Busywork begone. Meet Rovo, the AI teammate that actually knows your business. It helps teams find answers, make decisions, and automate work—securely and in context—through Rovo Search, Chat, Agents, and Studio. Powered by your organization’s knowledge, Rovo drafts specs in Confluence, creates work items in Jira, and even onboards new teammates. Work smarter. Get Rovo. |
|
Level up your career with these resources from our sponsors! |
|
|
BITS AND BYTES Stat: 20%. That’s how much sales of newly launched products increased for apparel brand Tecovas after it partnered with “global AI decisioning platform” invent.ai for inventory management, Revenue Brew reported. Quote: “Once you’ve got an agent sprawl, you potentially have a major cost problem…because you’re paying for a whole bunch of agents that you’re not necessarily using, or you are using but you’re not seeing the value from them or the value that you expected.”—Philip Carter, GM and group VP for IDC’s AI, data, and automation research practice, to IT Brew for a story about “agent washing.” Read: Why does so much new technology feel inspired by dystopian sci-fi movies? (The New York Times) Video star, you are: Wistia’s annual State of Video report breaks down data from 14 million videos, 100,000 businesses, and 1,300+ marketers to reveal what’s working. Get benchmarks and engagement trends by downloading the report.* *A message from our sponsor. |
|
|
SHARE THE BREW Share Tech Brew with your coworkers, acquire free Brew swag, and then make new friends as a result of your fresh Brew swag. We’re saying we’ll give you free stuff and more friends if you share a link. One link.  Your referral count: 0 Click to Share Or copy & paste your referral link to others: emergingtechbrew.com/r/?kid=073f0919 |
|
|
|
ADVERTISE // CAREERS // SHOP // FAQ Update your email preferences or unsubscribe . View our privacy policy . Copyright © 2025 Morning Brew Inc. All rights reserved. 22 W 19th St, 4th Floor, New York, NY 10011 |
|