Latest AI News
PhysicsWallah, Microsoft to Offer AI Courses Across Tier-2, 3 Cities
The programmes will be delivered through PW Skills and will integrate tools such as Microsoft Copilot, GitHub, and Microsoft Office 365.
View
How AI is Forcing Data Centers to Trade Uptime for Raw Power
As AI diversifies into training, inference, and hybrid workloads, operators are abandoning the one-size-fits-all approach for data centre infrastructure.
View
The Tech Players Powering India’s Predictive Police Push
Hyderabad Police’s AI deployment comes as city law enforcement agencies increasingly integrate digital intelligence into policing.
View
Indian IT Cannot Escape AI Cannibalisation
From TCS to Infosys and HCLTech, firms admit AI is eroding legacy revenues even as new growth builds.
View
The largest orbital compute cluster is open for business
For all the hype about data centers in space, there just aren’t very many GPUs up there. As that starts to change, the near-term business of orbital compute is starting to take shape. The largest compute cluster currently in orbit was launched by Canada’s Kepler Communications in January, and boasts about 40 Nvidia Orin edge processors onboard 10 operational satellites, all linked together by laser communications links. The company now has 18 customers, and announced its newest on Monday — Sophia Space, a startup that will test the software for itsunique orbital computeronboard Kepler’s constellation. Experts expect that we won’t see large-scale data centers like those envisioned by SpaceX or Blue Origin until the 2030s. The first step will be processing data that is collected in orbit to improve the capabilities of space-based sensors used by private companies and government agencies. Kepler doesn’t see itself as a data center company, but as infrastructure for applications in space, CEO Mina Mitry tells TechCrunch. It wants to be a layer that provides network services for other satellites in space, or drones and aircraft in the sky below. Sophia, on the other hand, is developing passively-cooled space computers that could solve one of thekey challengesfor large-scale data centers in orbit: keeping powerful processors from overheating without having to build and launch heavy, expensive active-cooling systems. In the new partnership, Sophia will upload its proprietary operating system to one of Kepler’s satellites and attempt to launch and configure it across six GPUs on two spacecraft. That sort of activity is table stakes in a terrestrial data center, and this is the first time it will be attempted in orbit. Making sure the software works in orbit will be a key de-risking exercise for Sophia ahead of its first planned satellite launch in late 2027. For Kepler, the partnership helps prove the utility of its network. Right now, it is carrying and processing data uploaded from the ground, or collected by hosted payloads on its own spacecraft. But as the sector matures, the company expects to start linking up with third-party satellites to provide networking and processing services. Mitry says satellite companies are now planning future assets around this model, pointing to the benefits of offloading processing for more power-hungry sensors, like synthetic aperture radar. The U.S. military is a key customer for that kind of work as it develops a new missile defense system predicated on satellites detecting and tracking threats. Kepler has already demonstrated a space-to-air laser link in a demo for the U.S. government. That kind of edge processing — dealing with data where it is collected for faster responsiveness — is where orbital data centers will initially prove their value. That vision sets Sophia and Kepler apart from established space companies like SpaceX andBlue Origin, or startups likeStarcloudandAetherfluxthat are raising significant capital to focus on large-scale data centers with data center-style processors. “Because we have the belief it’s more inference than training, we want more distributed GPUs that do inference, rather than one superpower GPU that has the training workload capacity,” Mitry told TechCrunch. “If this thing consumes kilowatts of power and you’re only running at 10% of the time, then that’s not super helpful. In our case, our GPUs are running 100% of the time.” And once these technologies are proven in orbit, well, anything can happen. Sophia CEO Rob DeMillo points out that Wisconsin adopted a ban on data center construction last week, something some lawmakers in Congress are also pushing. Anything that limits data centers on Earth is, in their eyes, making the space-based alternative more attractive. “There’s no more data centers in this country,” Demillo mused. “It’s gonna get weird from here.”
View
Accenture Invests in Replit, Partners to Scale Vibe-Coded Software
Replit offers a cloud-based platform that integrates coding environments, AI-assisted development, collaboration tools, and hosting.
View
Is India Ready for Drone Warfare? The Weakest Link is the Obvious
With the use of drones in the conflicts in Ukraine and West Asia, India is now stepping up its defence adoption.
View
Ramp’s AI Coworker Turns Employee Workflows Into Reusable Skills
“When one person on a team figures out a better workflow, everyone on that team gets it and gets more productive.”
View
New Research Finds Seven ‘Deadly’ Vulnerabilities in AI Benchmarks
A study from UC Berkeley showed how easy it is to game popular AI model evaluation tests.
View
Trump officials may be encouraging banks to test Anthropic’s Mythos model
Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell summoned bank executives for a meeting this week where they encouraged the executives to use Anthropic’s new Mythos model to detect vulnerabilities,according to Bloomberg. Indeed, while JPMorgan Chase was the only bank listed as one of the initial partner organizations with access to the model, Goldman Sachs, Citigroup, Bank of America, and Morgan Stanley are reportedly testing Mythos as well. Anthropicannounced the model this weekbut said it would be limiting access for now, in part because Mythos — despite not being trained specifically for cybersecurity — is too good at finding security vulnerabilities. (Others suggested this washypeor simplya smart enterprise sales strategy.) The report is particularly surprising since Anthropic iscurrently battling the Trump administration in courtover the Department of Defense’sdesignation of Anthropic as a supply-chain risk; that designation came after negotiations fell apart over the company’s efforts to limit how its AI models can be used by the government. Meanwhile,the Financial Times reportsthat U.K. financial regulators are also discussing the risk posed by Mythos.
View
Apple reportedly testing four designs for upcoming smart glasses
Apple plans to sell its first smart glasses in 2027, with a possible unveiling at the end of this year,according to Bloomberg’s Mark Gurman. Gurman has been reporting steadily on the evolution of the company’ssmart glasses strategy, but now he has more details about how they’ll look — he said Apple is testing four designs, and could ultimately launch with some or all of them. Those designs reportedly include a large rectangular frame, a slimmer rectangular frame (similar to the glasses worn by CEO Tim Cook), a larger oval or circular frame, and a smaller oval or circular frame. Apple is also considering different colors including black, ocean blue, and light brown. In some ways, these glasses are a step back from an ambitious plan that once called for Apple to launch a variety of mixed and augmented reality devices — a plan that already stumbled with product delays andthe lackluster reception of the Vision Pro. These glasses, meanwhile, sound closer to theMeta’s Ray-Ban glasses. They won’t have any displays, but will allow users to take photos and videos (Apple is reportedly oval camera lenses), answer phone calls, play music, and interact withthe long-promised Siri upgrade.
View
At the HumanX conference, everyone was talking about Claude
At the HumanX AI conference in San Francisco this week, thousands of techies descended upon the city’s Moscone Center, where discussion focused on the ways agentic AI is changing the business. Agents, which automate business and coding tasks, have begun to be deployed across industries — largely through enterprise and consumer-focused chatbots. Naturally, I wanted to know which chatbot was the most popular, and I consistently heard one name most often: Claude. Anthropic got shoutouts in many of the panels held throughout the week, but it also was a topic of discussion with the vendors I spoke to while perusing the convention room floor. The chatbot I didn’t hear a lot about? ChatGPT. One of the vendors I spoke to made a point of telling me that he and his team used Claude a lot, while he felt ChatGPT and OpenAI had gone downhill — or, as the internet likes to say, “fell off.” Lately, that does not appear to be a particularly unique take. Indeed, it’s not clear what will cure the perception that, despite a recent$122 billion funding roundand itsupcoming IPO, OpenAI has lost its footing—or, at the very least, seems increasingly unsure of what the next step is. Part of the problem may be a perception that the company lacks focus. Last month, OpenAI abandoned a number of long simmering side-quests (including its AIvideo generator Soraand a troubled plan to launcha “sexy” version of ChatGPT), locking in instead on the focuses of business and coding services. In the meantime, a number of developments, includinga recent New Yorker piecethat questioned whether the company’s CEO, Sam Altman, was trustworthy or not, have spurred a certainamount of negative buzzaround the company. The company’s work with the Trump administration hasn’t won it any friends either, nor has its decision to inject advertising into ChatGPT. During one of HumanX’s discussions, Sierra co-founder and CEO Bret Taylor (who is also the chairman of the board of OpenAI) defended Altman when asked by Alex Heath about the New Yorker profile. “I think Sam is one of the most visible leaders and executives in the world,” said Taylor. “If you want to seek out detractors for him, you’ll find them, and they’ll be very vocal about it,” he said, adding: “I think Sam’s remarkable. I think he’s a remarkable leader of AI, and I really trust his character as someone who’s worked with him.” The controversies and vacillations can make OpenAI’s seem reactive rather than strategic, as if it’s simply responding to events rather than shaping them. That said, when it comes to prominence and revenue, OpenAI and Anthropic are neck and neck — or at least, that’s how it looks, with some data suggesting thatAnthropic is catching up among business users. The Wall Street Journalrecently analyzedtheir finances, showing that the two companies were “the fastest-growing businesses in the history of tech.” In that sense, perhaps “falling off” for OpenAI just means it’s not the undisputed champ anymore. It has competition — which, in most industries, is normal. If anything, it remains clear that OpenAI is determined to do what it takes to remain dominant. This week, the company announceda new $100 subscription tier to ChatGPTwith substantially more access to Codex, its coding tool. The move seems clearly designed to spur broader use of tool while hopefully peeling users away from Claude Code. During a HumanX discussion with Bloomberg’s reporter Rachel Metz, OpenAI CTO of B2B applications Srinivas Narayanan noted how quickly the technological landscape has been changing. “We are in this incredible moment in technology, where every month, and sometimes every day, we are all looking forward to something new,” Narayanan said. Pointing to agentic coding as an example, he added, “We knew AI was going to impact software engineering, people have been using assistive coding over the last year, but even in just the last few months, the entire field has changed.” Agentic accomplishments may be a big focus of the tech community currently, since other applications for AI (creative uses, for example) haven’t really panned out yet. Still, the amount of work that companies have begun to offload onto their new little automated helpers is somewhat surprising—and, as Narayanan noted in his remarks, it has all happened in a relatively short period of time. In such an unpredictable environment, the future is still wide open.
View