Explore how Agentic AI moves beyond passive assistance to act more like a collaborator, using context, business processes, and existing IT environments to help organizations drive efficiency and new ways of working. The post How Agentic AI Is Changing Work appeared first on TechRepublic.

Amazon says its new ultra-fast delivery option brings groceries, household essentials, and other items to customers’ doors in 30 minutes.

AI-powered autonomous vehicles still have major hurdles to clear. Samuele Errico Piccarini/Unsplash, CC BY-SA Picture this: you’re driving on a mountain road, when you suddenly hit a thick patch of fog. You respond instinctively. Your vision sharpens, and you narrow your eyes to make out the shape of any oncoming cars. Human beings handle these quick changes very well, but if it were a self-driving car – at least one with a current artificial intelligence (AI) system behind the wheel – things could easily end in disaster. Today’s AI vision systems are extremely accurate when visibility is good. On a clear, sunny day a self-driving car can recognise pedestrians, road signs and other vehicles with precision. However, they are extremely vulnerable to environmental changes. If it rains, or gets dark or foggy, standard AI systems become blind, incapable of detecting obstacles that a human driver would spot with ease. Our research at the University of Valencia proposes a possible solution: instead of exposing AI models to millions of images of every possible road condition, we decided to imitate biology. But biologically speaking, why can humans see so well under such a wide range of conditions? Human vision: what we actually see – and don’t see – tells us a lot about consciousness The brain’s ‘volume control’ In our brains, neurons do not work alone. They use a truly fascinating form of adaptation that neuroscientists call divisive normalisation. To understand this (without getting into mathematics) we can picture it as an automated “volume control” system, with neurons working in a team. Let’s say one neuron is looking at a very dark area of the field of vision, such as a black car at night. The neighbouring neurons turn up the “volume” of this weak signal, amplifying the small details to make them more visible. If we look at a bright light, the same thing happens in reverse. The brain turns down the volume to prevent us from being dazzled. This mechanism is what allows us to adapt and see clearly in a very wide range of conditions. But in the search for speed and accuracy, modern AI systems have neglected this biological inspiration. AI systems and humans ‘see’ the world differently – and that’s why AI images look so garish AI in the driving simulator In our study, we processed images using some of the most widely used AI models, adding layers to simulate the brain’s “volume control” mechanism. In basic terms, we forced their neurons to communicate with one another and adapt to their environment, just as our own brains do. We wanted to see if imitating biology would make cars safer. To do this, we submitted both standard AI models and our brain-inspired modification to a series of tests. Using databases from real driving in European cities, night driving images from Switzerland, and several different virtual driving simulators, we were able to compare responses to difference levels of fog, darkness and light variation. The results showed that imitating our own brains worked. After being trained, the two types of AI models could drive perfectly well, but once fog and darkness came into the equation, the unmodified one began to fail. It lost the ability to distinguish cars from buildings, and even from the road itself. The AI system that was equipped with our brain-inspired mechanism, on the other hand, was robust. Even in fog or complete darkness, it performed more than 20% better than its unaltered counterpart. We analysed, from the inside, how this new system perceived the world and found that it was doing exactly what we expected. It was capturing and enhancing the details of vehicles hidden in the fog that would otherwise be invisible. As a result, its performance became more stable in the face of changing weather conditions. The next generation of driverless cars will have to think about what’s on the road, not just see it Learning from nature Getting society as a whole to trust AI poses major challenges, and the safety of passengers and pedestrians in self-driving cars is a major aspect of this. It is not enough for smart systems to work under ideal conditions. We need them to be completely safe in the real world, and to safeguard the lives of all road users in all weather conditions. Our research shows that the key to making artificial intelligence safer, more robust and more adaptable may be closer than it seems. There is no need for more powerful computers or vastly greater amounts of data. Sometimes, all we need is to look at the millions of years of evolution that have shaped our own brains. In many cases, nature has already solved some of the problems that artificial intelligence faces today. We just need to learn from it. A weekly e-mail in English featuring expertise from scholars and researchers. It provides an introduction to the diversity of research coming out of the continent and considers some of the key issues facing European countries. Get the newsletter! Pablo Hernández Cámara receives funding from the Spanish Ministry of Science, Innovation and Universities.

Discussion | Link

PLUS: Build a YouTube research bot in 15 minutes

Laserfiche has announced the release of AI agents that can help perform tasks through natural language prompts. Intelligent assistants follow Laserfiche’s integrated security rules and compliance requirements, helping ensure all sensitive data remains protected. Karl Chan, CEO of Laserfiche, said, “The introduction of AI Agents to content management signals a change in how we handle […] The post Laserfiche unveils AI agents for natural language workflows appeared first on AI News.

Discussion | Link

Ramanathan said his Ph.D taught him to value skepticism and verification. Courtesy of Anand Ramanathan Anand Ramanthan ignored advice to major in computer science, and instead studied physics to Ph.D level. He pivoted to data science and said the skills he gained during his Ph.D serve him in Big Tech. But it can be hard to explain how Ph.D skills translate to a corporate job, Ramanathan said. This as-told-to essay is based on a conversation with Anand Ramanathan, who works at Meta and lives in New Jersey. The following has been edited for length and clarity. When I was choosing my undergraduate degree in the early 2000s, people told me to pick computer science, the hot field of the time. Instead, I studied physics, all the way to getting my Ph.D in 2011, because I really enjoyed it. In 2012, after working at NASA for roughly a year, I was thinking about what was next, and I looked into fields like defense contracting. My cousin, with good intentions, advised me to downplay my Ph.D in applications, by not bolding it on my résumé and placing it lower down the page than my other experience. It shocked me to the core. I'd worked so hard and learned a lot of skills, and felt my Ph.D had to be useful somewhere. In the end, I stayed with NASA for several more years before pivoting into the data science industry. Now, I work at Meta. I'm really glad I got my Ph.D and I've come to understand how some of the skills it helped me develop are valuable in Big Tech. I wanted to do something with real-world applications Ramanthan's time at NASA led him to the field of data science. Courtesy of Anand Ramanthan My Ph.D at the University of Maryland focused on cold atom physics and involved a lot of research and lab duties, which made it feel more like I was working full-time rather than studying. When I finished, I wanted to do something with real-world applications and landed a postdoctoral fellowship at NASA. I spent six years there as a contracted research scientist, where I worked with a lot of data, which I really enjoyed. This became a bridge for me into the field of data science, and I also did a few related courses to upskill. I saw how my Ph.D and research experience could be transferable to the data science industry, and between 2017 and 2024, I worked as a data analyst and scientist at Capital One, and as a data scientist at Audible. Since July 2024, I've worked at Meta, where I use data to improve products. My Ph.D taught me to communicate complex ideas, and be skeptical and curious There are a lot of talented people around me at Meta, and although I don't use a lot of the physics knowledge I gained from my Ph.D today, I think it sets me apart in some ways. First, it taught me to value skepticism and verification. I was so enthusiastic about my findings, but my Ph.D supervisors were very skeptical and careful about subscribing to anything I presented to them. Now, after I've checked my work, I'll often ask colleagues or domain experts to review it before circulating it broadly, especially when a result is surprising or business-critical. During my Ph.D., I also spent many, many hours writing papers, including my thesis that conveyed the technical depth of my work in a way that others could understand. I used these skills when writing summary documents for senior leaders at Audible. And there's no promise of a $500k job at the end of a Ph.D, so you're motivated by your curiosity, and learn to ask a lot of questions, which has helped me navigate the fast-changing career landscape. When AI LLMs first came out, I played around with them and took a couple of courses. I now feel like I know the basics of how LLMs behave thanks to my ability to learn and adapt. I've had to go the extra mile to show how my Ph.D skills translate I'm grateful for what I learned from my Ph.D, but because it can be hard to explain how it can apply to a job, I feel like I've had to go the extra mile to communicate what skills I bring to the table. I wouldn't advise doing a science Ph.D as a career move unless you like the field and love what you're doing. But if you have done one, know that you do carry value. You may just have to put in a little more effort to translate that value when applying for jobs. Do you have a story to share about breaking into Big Tech? Contact this reporter at ccheong@businessinsider.com Read the original article on Business Insider

A new paper out of arXiv this week describes an AI system that builds, improves, and deploys its own specialist agents. Here is what that actually means for engineers and technical teams.

Getty images; Tyler Le/BI Like a lot of guys of the era, Eli Cohen had a great idea for an app back in 2010. But unlike many of his peers, whose musings were confined to bar hangouts and bad dates, he actually invested about $20,000 of his own money to try to bring the project to life. Unfortunately, his gumption only got him so far. Back then, "building software was painfully hard," he says. He burned through his budget well before it got off the ground. What bugs him isn't that the endeavor failed, it's that a similar concept — the online education platform Udemy — launched that same year and went on to IPO at a $4 billion valuation in 2021. "I assume they had more resources, better developers, of course, luck,," Cohen says, emphasizing that he's not bitter about it. The main lesson he took from the experience is that sometimes the difference between success and failure is not the idea but the ability to execute. Cohen's entrepreneurial journey didn't end there. The 45-year-old, who lives in Israel, is now building a meditation app he's dubbed MediTailor. He and his brother used AI to create in weeks what would have taken a development team 18 months not long ago. "For the first time in 20 years, the wall between having an idea and actually building it has come down," he says. Cohen is verbalizing a sensation familiar to many people in the AI era. Armed with a Mac Mini and a Claude subscription, starry-eyed entrepreneurs can vibe code their way into the App Store with relative ease. Gone are the days of that stereotypical guy who insists he has a killer idea for an app… if only he could bring it to fruition. The former Silicon Valley punchline is now a credible threat: New app releases worldwide so far this year have more than doubled from last year. The tech industry's technical moat may have shrunk to a puddle, but just because any guy (or gal) with an idea for the next "Uber for whatever" can make it happen doesn't mean they'll create the next internet giant. Execution still matters, and the vast majority of apps still fail. A more saturated market also makes it harder to stand out, and most ideas are not hugely scalable, easily marketable, or particularly unique. When I point out to Cohen that there are already countless meditation apps, he tells me he's well aware. "Maybe I will not succeed, but at least I have the tools," he says. So does everyone else. To be sure, it's not the case that literally anybody can vibe code an app these days — many people don't have the time, energy, or desire to figure out how to use OpenClaw or Claude Code. "The barrier isn't zero right now," says Rebecca Kaden, a managing partner at Union Square Ventures. While vibe coding allows for a new level of "efficient entrepreneurship," we're not at true democratization of technology. Vibe coding tools aren't plug-and-play yet — it's not a "no brainer" for any random consumer to wake up one morning and start doing it. The former Silicon Valley punchline is now a credible threat. For those with the will, however, the way is more accessible than ever. That's leading to an avalanche of apps: Data from market intelligence firm Appfigures shows that 414,000 new iOS and Android apps were released in the first quarter of 2026, up 115% from the same period in 2025. Also during that same period, three times as many apps were updated compared to the prior year, signaling rapid shipping, experimentation, and churn. It's super hard to get attention, though. A mere 118 new apps achieved what Appfigures calls "high-traction" status — meaning more than 50,000 downloads in the US — during the first three months of the year, a measly 0.02% hit rate. It's also a relatively tiny increase from the 94 apps that did so in 2025. "The app stores may be entering a higher-noise era: more experiments, more fast launches, and a smaller share that turn into meaningful businesses," says Ariel Michaeli, the CEO and cofounder of Appfigures. There's a famous '90s episode of "South Park" where the crew meets a group of "underpant gnomes" who develop a questionable three-phase business plan. They have a handy chart that shows three steps to riches: Phase one is to collect underpants. Phase three is profit. And phase two is a giant "?." The vibe coding environment is leading to a bit of "underpants gnome logic," explains Charity Majors, an operations and database engineer and the CTO and cofounder of Honeycomb.io. People can collect the underpants — as in code the thing — but there's a giant question mark before arriving at profits. Code that's spun up, more or less for free, still requires modeling and maintenance. Plenty of people could vibe code a Slack clone, but the hard part of building Slack wasn't writing the code in the first place, Majors says. It was thinking up the design that made it easy to use, intuitive, and scalable to billions of people. "Nobody wants to launch Slack every day and see their buttons move around," Majors says. One risk of vibe coding an app without any technical knowledge is that people have no idea how to fix something when it breaks — it creates a cottage industry for software engineers who can address bugs or manage security. The idea's not great because you don't understand how many millions of people need to agree with you. Building and maintaining are not the only gaps between "genius idea" and "sustainable company." There's also — very importantly — marketing and reach. Getting into Google Play or the App Store doesn't mean anyone will ever see or use your app. Kate Minogue, an AI advisor who works as a fractional chief product officer for consumer tech startups, says distribution remains the hardest problem to solve, and the proliferation of apps makes it worse, not better. "There's always an Uber-for or Duolingo-for, and you're like, 'OK, well, it exists, or there isn't actually a market for it,'" Minogue says. "A lot of that time it's like, 'Well, the idea's not great because you don't understand how many millions of people need to agree with you that this is a great idea for it to make money.'" A bunch of people sitting around an apartment in a high-income city is not exactly a focus group for what a massive amount of the population wants or needs. Your awesome business idea may not sound so awesome to everyday people — or even to yourself in the cold light of day. Wannabe entrepreneurs do what they can to optimize their performance in app stores, search engines, and LLM responses in order to try to climb the rankings, or pray they make a viral hit. If that doesn't work, they've got to go to paid marketing and advertising. "When I talk to companies that are building an app, I try to prepare them for having to spend hundreds of thousands of dollars to get scaled distribution of their apps," Minogue says. Some businesses are navigating this by taking a "studio-based" development approach where they release five, 10 different fitness or productivity apps at once and test in real time whether any of them stick. They don't necessarily need a banger — they just want one or two to get enough traction to generate some cash. In the meantime, the orphans contribute to the growing mass of apps. Amjad Masad, the CEO of vibe-coding platform Replit, says it's a good thing for more "wantrepreneurs" to be able to give their ideas a go — democratization, or however close this is to it, will inevitably mean variations of quality. Not everyone's setting out to make a gazillion-dollar product, either. Some people are vibe coding tools and widgets to make it easier to get by at work. Others are making apps that, while not enough to support a full-blown corporation, are enough to support their day-to-day expenses. "The cool thing about it is that those apps, a lot of them don't have to be venture-scale, multi-billion-dollar apps," Masad says. "They can still make a lot of money for the founders, and they can serve a need." For some of the Silicon Valley old guard, this is starting to feel like an identity crisis: If a random guy can vibe code a functional product in a weekend, what was the point of fighting over exclusive talent and battling competitors to the death over the past 20 years? Sure, the big players who've already made their millions — and billions and trillions — are thrilled about the prospect of cutting costs and speeding up via AI, but what about the people who've been slogging to break in? Kylan Gibbs, the cofounder and CEO of Inworld AI, who previously worked for Google DeepMind, tells me a lot of Silicon Valley natives are starting to "freak out" over recent AI developments. The fear is that either apps are now so easy to build that 100 competitors are about to pop up, or that OpenAI or Anthropic can launch a similar product that will squash their budding firms overnight. Say you've been working on an awesome fitness app for five years, grinding to get downloads and perfect the offer. Suddenly, the app store is flooded with options that, to potential users, are indistinguishable — even if they're actually buggy or clunky. At the same time, there's a sword hanging above your head: What's to stop one of the big players from making its own fitness add-on? They're already slurping up your customers anyway, because who needs a special app for workout advice when they can ask Gemini to create a plan for free? For some of the Silicon Valley old guard, this is starting to feel like an identity crisis. "I've literally in the last few weeks had a good handful of friends and colleagues who are other founders say, 'Yeah, I'm just basically giving up,'" Gibbs says. "You think about this group of people who have felt so special for so long, and then to have that ripped away." Terrence Johnson, a software engineer in Colorado, is looking at the bright side of the everybody-can-code attitude: It means fewer people bugging him to cook up their projects. He recently dealt with a particularly pushy individual who was trying to enlist him into making some sort of car dealership management system without any concept of what it would entail or how long it would take. He realized the person was getting their ideas from AI in the first place, so Johnson nudged them toward other tools that would let them do some cursory stuff on their own. "It functioned as enough of a distraction that they left me alone," Johnson says. "Empowering people to try and get their ideas out is good, but they just don't quite grasp what it all entails." New crops of entrepreneurs may be in for a rude awakening on the lessons many startup veterans are familiar with. It's next to impossible to build a blockbuster. Most endeavors flop. Making money has to, at some point, turn from a future problem into a now problem. Viaano Spruyt, who lives in Singapore, is excited that vibe coding has allowed him to turn his online community focused on mental health — which formerly lived on Discord — into an app called Huddle. The new app allows them to add more features and functionality, and he and his cofounder have even secured some investor capital. They're pivoting their concept from mental health to catering to fandoms, but the profit part of the plan remains a question mark, and they're relying on word of mouth for distribution. "What's becoming harder is getting people to care," he says. Spruyt and his cofounder are learning what a lot of vibe-preneurs are beginning to realize: Business basics are extra important, and excellence in sales, marketing, and logistics is a way to get ahead — or at least manage to claw your way out of the crowded trenches. Relatively speaking, there are worse problems than having 900 too many productivity apps in the App Store and forcing the Silicon Valley elite to try harder. A builder economy is a fun one to be in. And hey, if it shuts up the killer app idea guy who's talking your ear off over at a bar holding an IPA, it's a win for everybody. Emily Stewart is a senior correspondent at Business Insider, writing about business and the economy. Read the original article on Business Insider

Discussion | Link

Discussion | Link

Discussion | Link

Late-night hosts said Stephen Colbert could host their shows whenever he wants to. Danny Moloshok/AP The big names of late-night got together during one of Stephen Colbert's last episodes. Jimmy Kimmel and Jimmy Fallon joked that Colbert was welcome to host their shows. CBS said last year that it would pull Colbert's show, and May 21 would be his last taping. Late-night hosts joked about giving Stephen Colbert a gig on their shows once his is canceled. Colbert brought five of the biggest names in late-night under his roof for his Monday taping of "The Late Show with Stephen Colbert." Jimmy Kimmel, Jimmy Fallon, John Oliver, and Seth Meyers joined Colbert to talk about the future of late-night, getting slammed by President Donald Trump, and Colbert's looming exit from CBS on May 21. Colbert said it was a "bummer" to end the journey because he wanted to keep the late-night show going with the crew. "Now I can be a guest on all your shows," he said. Kimmel immediately jumped in, saying "You can host my show." Colbert asked if he could host "Jimmy Kimmel Live!" in the summertime, and Kimmel replied that he could be there all year. Fallon also told Colbert that he was welcome on his show "anytime you want to go on." When Colbert asked him when it would be convenient for Fallon to let him substitute, Fallon jokingly dodged the question, saying, "Enjoy a little time off." CBS announced in July that "The Late Show with Stephen Colbert" would be canceled in May, after more than three decades on air. The network said it was a purely financial decision. News of the cancellation came shortly after Colbert made a joke about CBS paying a $16 million settlement to Trump over a lawsuit that accused CBS of "deceptive editing" in a "60 Minutes" episode. Colbert's exit comes as late-night has faced a stressful year. "Jimmy Kimmel Live!" was suspended for about a week in September after Kimmel's comments on political activist Charlie Kirk's shooting angered the Trump administration. His show was reinstated after support poured in from Hollywood and politicians. Trump and first lady Melania Trump slammed Kimmel again in April for a joke he made about Melania having a "glow like an expectant widow," and called for his suspension again. Read the original article on Business Insider

Japanese snackmaker Calbee is simplifying packaging as a naphtha supply squeeze hits printing ink. Calbee A popular potato chip maker in Japan is turning its snack bags black and white due to naphtha shortages. Prices of naphtha, a raw material for printing ink, have soared as the war in Iran is squeezing supply. The supply squeeze is hitting companies across Japan, including toiletmaker Toto. First, fuel prices surged. Now, even potato chips aren't safe from the global petrochemical squeeze. In Japan, snack giant Calbee — known for its popular potato chips — is switching some colored snack packaging to black and white after oil shortages linked to the Iran war squeezed global ink supplies. "This measure is intended to help maintain a stable supply of products," Calbee said in a press release on Tuesday. The company attributed the temporary move to "supply instability affecting certain raw materials." The culprit is naphtha, an oil-derived chemical used to make solvents and resins for printing ink. Naphtha is also a key ingredient in plastics, packaging material, and adhesives. Prices for the material have surged amid the conflict in Iran. Calbee said 14 product types will appear in monochrome packaging from May 25. A photo of the new packaging from Calbee showed a note on the left-hand side explaining that the monochrome design was intended to conserve petroleum-derived raw materials. "The change applies only to a limited number of products and will not affect product quality," Calbee said in the press release. The move underscores Japan's exposure to disruptions in naphtha supplies. Japan relies heavily on imported naphtha, making it especially vulnerable to supply disruptions. In recent weeks, companies including toilet maker Toto and Panasonic have warned of delivery disruptions and price hikes tied to naphtha-based materials. In the food industry, Calbee is not the only company facing naphtha-related pressure. Earlier this month, Itoham Yonekyu — a Japanese maker of processed meat and precooked food — said it is planning to reduce the number of colors used in product packaging to cut costs. Japanese officials have tried to calm fears of a broader shortage. On Tuesday, Kei Sato, the country's deputy chief cabinet secretary, said there are no immediate problems with Japan's naphtha supply. He added that while 40% of the country's naphtha supply comes from the Middle East, an equal amount is produced domestically. Read the original article on Business Insider

Mario Lopez says he and his wife, Courtney, "never stopped dating." Vivien Killilea/Getty Images for Calamigos Ranch Mario Lopez says he and his wife, Courtney, still do this one thing to help keep the spark alive. "We never stopped dating — that's important," Lopez said. The couple has been married since 2012 and shares three kids. Mario Lopez, 52, says he and his wife still make time for one thing that keeps them close. Speaking to People at an event held in Malibu on May 1, he shared that they still go on regular dates and trips together. "We try to carve out certain days," Lopez said in the interview, which was published on Sunday. "This little event is a little getaway weekend within town, which is great, so we get to be back on Sunday. Those are important. We never stopped dating — that's important." Lopez shares three kids with his wife, whom he married in 2012. "It's funny because we like to go to Vegas a lot," Lopez said. "We go for the fights, we'll go for shows and stuff, do the wine trips. This is great. Catch up on sleep." Their ability to genuinely enjoy each other's company has remained central to their relationship. "It just gets better and better, and I'm not trying to be corny, but it does. We just love each other. We have fun. We laugh," his wife, Courtney Lopez, added. This isn't the first time the "Saved by the Bell" actor has talked about his marriage. In a December interview with People, Lopez said they avoid spending "more than two weeks without seeing each other," don't go to bed angry, and prioritize date nights. Lopez is one of several celebrities who have spoken about the importance of intentionally carving out time with their partner. In September 2024, Gordon Ramsay's wife said she and her husband make a point of dressing up for their date nights. "That's where we always have our best chats. We go out to dinner and we talk about ourselves, we talk about each other, we talk about the kids, but it's not just about the kids," she said. In May 2025, Kelsey Grammar said the key to his marriage is putting in constant effort. "I always say, 'Love is a contact sport,'" Grammer said. "Once in a while you got to back it up with some action." Read the original article on Business Insider

Mati Staniszewski says adding an engineer to each team can upskill it. Big Event Media/Getty Images for HumanX Conference ElevenLabs CEO Mati Staniszewski said he's adding engineers to non-technical teams. He said that non-technical teams around the company are already building tools for their use cases. The company has a micro-team structure for end-to-end ownership. There's one thing you need before you can vibe code everything: An engineer to show you how. At a talk at Sequoia Capital released last week, ElevenLabs CEO Mati Staniszewski said that he is adding an engineer to every non-technical team inside the voice AI company. "Our people team, our go-to-market team, our legal team will have an engineer in that team that helps to build automation and upskill, uplevel the rest of the people," he said. "Recently, that really helps because everybody will be vibe coding and coding a lot." Staniszewski, who cofounded ElevenLabs in 2022, said that non-technical teams around the company are already building tools. "There's just so much incredible work you can do by having that, whether that's the scraping on the hiring and recruiting front or analyzing what worked in the past to improve in the future," he said. The company "recently introduced a scoring system for those on the go-to-market or the sales side." He said the scoring system tool has saved many negotiations with the sales team, which used to ask him about provisions they could give a potential customer. In February, ElevenLabs announced that it raised $500 million in a Series D round, bringing its valuation to $11 billion. The startup, which was founded in London, had 350 employees as of November, according to investor Andreessen Horowitz. End-to-end ownership The company seems to follow an increasingly popular organizational model in which teams are expected to take on greater ownership and be self-sufficient. This means engineering teams are responsible for everything from product design to marketing what they've built, and non-technical teams are expected to create their own tools to improve workflows. ElevenLabs' employees are divided into about 20 micro-teams of five to 10 people each, Staniszewski said in a September interview. Each of these teams owns a product area, such as a studio interface or voice agents, to move fast, he said. Execs at AI coding startups, including Cognition and General Catalyst-backed Kilo, told Business Insider that they view ownership as a must-have for current and potential employees. "You see a problem, solve the problem," said Emily Cohen, who heads people and operations at Cognition, in an interview last month. "We're not the classic company that's just like, 'Oh, well that's that team's job.'" Read the original article on Business Insider

Financial Times: Sources: some Amazon employees are using in-house OpenClaw-like tool MeshClaw for unnecessary tasks to inflate AI token use after Amazon set weekly AI targets  —  In-house MeshClaw tool enables employees to delegate jobs to AI agents and climb company's AI leaderboard

Right now, every AI model you've ever used works the same way. You talk, it listens. It responds, you listen. Thinking Machines is trying to change that by building a model that processes your input and generates a response at the same time, so it's more like a phone call than a text chain.

Wall Street Journal: The US House Oversight Committee launches a probe into potential conflicts in Sam Altman's personal investments; letter: several GOP AGs call for an SEC review  —  Republican-led House Oversight Committee says it is investigating, and six GOP state attorneys general are calling for SEC review after WSJ article