The company has set aside an unusually high number of shares for retail investors. Still, experts say, you’re just getting the crumbs.

Discussion | Link

Kristi Edleson is the chief of staff at Yutori Courtesy of Kristi Edleson Kristi Edleson, a startup chief of staff, uses an AI version of herself to help with her workload. She said the hardest part is knowing when to do the work herself or outsource it to her AI agent. Despite relying on her AI chief of staff for numerous tasks, she refuses to let it handle finances. This as-told-to essay is based on a conversation with Kristi Edleson, a 34-year-old chief of staff at Yutori, based in San Francisco. The following has been edited for length and clarity. I work at a startup as the chief of staff, and my company's goal has always been to build an AI chief of staff. It's funny that when I took this role, I was signing up for that. I wasn't nervous about it, though; I actually loved the irony. A chief of staff's role involves a lot of context switching, and I believe the AI product my company built focuses more on augmenting my skills than fully allowing me to hand things off. The product launched recently, but I don't see the agent as a replacement for my full role. I think finding the balance of what lives in your head versus what you're willing to outsource is the hardest part of working with AI agents all day. Business Insider wants to talk to founders of AI-powered companies with fewer than 10 workers, and employees working alongside AI agents, to understand what the "Tiny Teams" era really feels like — the wins, the fears, and the human skills that stand out. Share your story by filling out this quick form. I'm the only non-tech person at the AI company I work for I'd been working in recruiting for the past 10 years, and I wanted to make a pivot into something more operational. This role happened by chance spring 2025. I knew the founders from when I was a recruiter at Meta, and they had a catch-all kind of position they wanted to fill at their startup. I don't think we had even named the job chief of staff yet. As conversations evolved, that's the title we landed on. I never saw chief of staff as my forever role, but this job felt like one of the most high-impact operational roles I could get. Who better to give product feedback than someone who is living those tasks daily? I'm the only non-technical person on the staff, and every day can be different. My job is basically anything the founders don't have time for. That can be daily operations, finance, HR, compliance, conversations with different vendors, negotiating contracts, and reading contracts. I give the AI chief of staff agent my manual work I have our AI chief of staff connected to all my work tools, like email, Slack, Linear, Granola, etc. So it can pull in all of that context and pick up the things that I forgot about. The chief of staff agent also pulls from my calendar to alert me of what else is on our plate. From there, I decide which tasks I want to hand off. Before AI agents, I'd see my task list and know I had to do everything on it. Now I ask myself: what items can I hand off to AI so I can find even more meaningful problems to solve? That means I also need to go find those problems. One of the things that I like to do in the mornings is switch into voice mode, do a brain dump, and unload everything I'm thinking about in any order of chaos. I'm not someone who's super into having set prompts. That's why I love where agents are at, at this point. I don't have to be perfect with my prompts and can ask whatever is on my mind in a very imperfect way, and it will still work. I like to challenge myself anytime there's something that feels a little bit manual. How can I have an agent do that on my behalf? That could be anything from drafting emails or Slack messages to having ChatGPT help design our office swag. I like having reach over what the agent is doing, so it will draft items, like an email, but I approve most of them before it hits send. There haven't been any major mistakes from my agent, but I have this step in there to safeguard against mistakes the agent could make. AI helps me negotiate vendor contracts and not ask dumb questions One of the core parts of my job is negotiating with vendors for different kinds of technical infrastructure tools we need to run our models, like compute or inference. I'm non-technical, so I use AI to help me not ask silly questions in these meetings or slow down our timeline. About 30 minutes before a call, I ask my agent, "What's the latest? What's the sentiment from the team on this tool?" It has my context for how we use certain tools, as well as conversations other teams are having about those tools in a public channel. I go back and forth with it and sometimes ask questions that I would be embarrassed to ask the team. Then I'm able to show up in vendor conversations, have authority, and get the things we need. I don't think AI can replace me for financial tasks and strategy mapping I don't use the chief of staff AI agent for financial tasks right now, probably because I'm risk-averse, and maybe I'm overcautious. I think finances are probably where I, and a lot of people, draw the line. A bad email is annoying but easy to apologize for. You can't un-spend $500. From a personal perspective of keeping myself sharp, I still like to write and handle a lot of tasks myself, then have agents gut-check them, especially anything strategic. I like to have a mental map of everything I'm working on, and AI sometimes works faster than my brain can keep up. I'm working on how to keep all that information somewhere in my mind, even if it's getting less attention overall. There's a lot of buzz about full automation or the end of certain roles, but I've been happily surprised by how much more empowered I feel as a worker now because I have access to basically the world's most patient tutor. Do you have a similar story to share? If so, please reach out to the reporter at aapplegate@businessinsider.com. Read the original article on Business Insider

Kevin O'Leary's company, O'Leary Digital, and West GenCo LLC are developing a massive data center campus in Utah. Michael Tran/AFP via Getty Images Kevin O'Leary is backing a massive data center campus in Box Elder County, Utah. The project has faced resistance from residents and local officials. O'Leary, however, says he doesn't "tap out" when projects become contentious. Kevin O'Leary is not one to give up. The "Shark Tank" investor, who also this year took a spin as the villainous businessman in "Marty Supreme," says that despite all the local resistance to his massive Utah data center, he's holding firm. "I don't tap out. I don't even know what that means. That's never going to happen with me," O'Leary told Business Insider. "Everything I do has challenges. Every deal, every project — it's never easy." Building a hyperscale data center, the kind that powers the products made by AI companies like OpenAI, Anthropic, Google, or Meta, is not easy these days. Numerous proposed data centers across the country have faced opposition, and dozens of towns and cities have either passed moratoriums, outright banned them, or are considering it. O'Leary's proposed data center in Box Elder County, a rural area of about 60,000 people near the rapidly diminishing Great Salt Lake, is perhaps the most well known, thanks in part to O'Leary's celebrity status. As soon as details began to emerge publicly about his data center campus, called the Stratos Project, Utah residents began organizing against it. The opposition crowded local planning meetings, circulated petitions, and recently filed a lawsuit to stop it. The backlash got the attention of the state's governor and senate president, who both initially supported the idea. The two have, in recent weeks, issued new parameters around the development, making it harder for O'Leary to realize his lofty plans. None of this has deterred O'Leary. "There are bumps in the road. There's litigation. There's misinformation. There's propaganda. You name it. But if you know what you're doing is right, you make it, you get there," O'Leary said. "Persevere." Utahns opposing the Stratos Project protest outside the Utah State Capitol in May. Natalie Behring/Getty Images Sometimes called Wonder Valley, a reference to O'Leary's chosen nickname, Mr. Wonderful, the data center proposal initially spanned 40,000 acres. When fully online, that plan estimated the campus would have up to 9 gigawatts of power generation capacity. That's gigantic. Most large-scale AI data center proposals are at their largest 1 gigawatt. Many of them are closer to 300,000-500,000 megawatts. O'Leary Digital and West GenCo LLC are developing the project. In May, Box Elder County commissioners authorized a project area for the campus, eliciting furor among residents who oppose it. In addition to its vast size, critics are worried the data center will have an adverse impact on local wildlife, water resources, and air quality. They've also castigated local officials and developers over what they consider a lack of transparency. O'Leary says all those concerns are not based in reality. He, without providing evidence, blamed a Chinese disinformation campaign for stoking anxiety. The United States and China are in a race to lead the world in artificial intelligence. Data centers are a key leg of that race. "For me, all roads lead back to China," O'Leary said. O'Leary says the data center will create jobs and economic growth, and that the technology underpinning data centers — like cooling techniques — has advanced to the point that their environmental impact is negligible. Despite O'Leary's efforts to change public perception of data centers, Utah's government has responded to local concerns. Gov. Spencer J. Cox signed an executive order at the end of May that established a "higher bar for data center development in Utah." Utah Senate President J. Stuart Adam, meanwhile, called for a 75% reduction in the project area and other commitments in a letter sent to Digital O'Leary this month. O'Leary agreed to scale back the campus and adhere to other requests. "We're only going to build 1.4 gigawatts to start, which is sort of like a test kitchen," O'Leary told Business Insider. The project was always planned to expand in phases. "What's good about doing something small to start is you prove out your model, and you show the new technology at work, which is remarkably better than technology was in data centers 20 years ago," he said. Although new parameters have slowed down O'Leary's plans, Americans will likely appreciate the more measured pace. A June Reuters and Ipsos poll said only 1 in 3 Americans approve of the "fast pace" of AI data center construction. O'Leary, however, says that a fast pace is necessary as America strives to lead the AI race. "Ask yourself, can we build anything in America anymore, or is it always going to be someone telling us we can't?" O'Leary said. "I think those times are over — well, they have to be — because we're in a global competition, in an economic competition, in a military competition, and certainly a technological competition." Read the original article on Business Insider

Nvidia CEO Jensen Huang shows off the company's Blackwell system on stage. Ann Wang/REUTERS There's been a lot of hand-wringing lately about AI token costs. There could be relief coming later this year. A wave of new technology may cause token prices to plummet. It may already be starting to happen. I had lunch with the CEO of an AI infrastructure company recently. I can't tell you their name, but they said something that really caught my attention: There will be a crop of new AI models later this year that will be a lot better and more efficient. This will likely make AI tokens more abundant and radically cheaper. (Tokens are the basic units models use to process information, and the standard way AI use is measured and priced). Hand-wringing about tokenmaxxing could die down. Or, users could go on another bender and burn even more tokens with abandon. Either way, the price of tokens is probably about to plummet. This is why we already see some AI model providers slashing prices, and other players talking about doing so. OpenAI CEO Sam Altman recently said AI costs had become a huge issue, adding that the startup will have "a lot of ways we can help people get more value for less spend." This trend may already be showing up in the data. A closely watched token spending index run by Silicon Data peaked at around 2.06 in late May and fell to 1.75 as of June 10. Carmen Li, the CEO of Silicon Data, told me this could mean token prices are dropping across many AI models. Blackwell finally emerges The main force driving token prices lower is a new wave of technology that's sweeping through AI data centers. Nvidia's Blackwell GPUs are being installed in huge volumes right now. By the second half of this year, these systems, which are really supercomputers rather than chips, will be operating at scale, helping AI labs train new models and run them more efficiently. These systems took a while to install properly, partly because they needed to be water-cooled and required other gnarly new data center setups. But the payoff could be huge. 50 x more, 35 x cheaper SemiAnalysis, a respected AI research firm, compared Nvidia's top Blackwell system, the GB 300 NVL72, to Nvidia's previous system, called the Hopper HGX 200. With the older system, each GPU generated 90 tokens per second, while the new Blackwell system generated 6,000. That's 65 times more. These systems consume massive amounts of electricity, and the newer Blackwell offerings use even more. So SemiAnalysis also looked at how many tokens each system generated per megawatt. On this measure, Hopper churned out 54,000 tokens per second, while Blackwell generated 2.8 million. 50 times more. Electricity prices are rising, due to all these energy-sipping AI data centers. So these days, GPU systems are assessed based on how much it costs to generate one million tokens. SemiAnalysis tested this, too, and found that the older Hopper system cost $4.20 for every million tokens. The Blackwell system cost 12 cents. That's 35 times cheaper. Again, new AI models will be increasingly trained and run on these new Blackwell systems as 2026 progresses. This is very likely to produce a massive increase in the number of cheaply-generated tokens. This is why AI model providers will probably slash token prices: Because they can. Sign up for BI's Tech Memo newsletter here. Reach out to me via email at abarr@businessinsider.com. Read the original article on Business Insider

MIT Technology Review Explains: Let our writers untangle the complex, messy world of science and technology to help you understand what’s coming next. You can read more from the series here. Your brain lives in the dark space of your skull. Yet it knows when the wind lifts the hairs on your skin, when your heart is…

Earlier this week, Life Biosciences, a biotech company focused on reversing age-related diseases, announced that it had dosed its first volunteer. A person with glaucoma has had an experimental treatment injected straight into their eyeball. The idea is to try to treat the disease—which can cause vision loss—by regenerating healthy nerves in the eye. But…

PLUS: Use this X + Openclaw setup to write viral social content

The rocket maker debuts on Nasdaq today under a wave of criticism about Musk’s near-absolute control. It’s how the company has worked from the start.

Customer segmentation is the practice of dividing an existing customer base into smaller...

Discussion | Link

Discussion | Link

Discussion | Link

New York Times: How Chinese manufacturers are dominating the humanoid robot supply chain, even as the industry struggles to find a purpose for such robots  —  China's lead in humanoid robots was evident at last month's Humanoids Summit, a robotics conference in Tokyo.Videos by Qilai Shen and Hiroko Masuike

The AI Industry Said Safety and Capability Trade Off. Claude Fable 5 Disagrees  The benchmark tables … Claude Fable 5: The Model Everyone Said Couldn’t Exist Is Now Available to EveryoneRead more The post Claude Fable 5: The Model Everyone Said Couldn’t Exist Is Now Available to Everyone appeared first on .

Ivan Mehta / TechCrunch: Hyderabad, India-based Equal AI, which makes an eponymous AI-powered call screening app, raised a $30M Series B led by Prosus Ventures and Tomales Bay Capital  —  In India, consumers receive a lot of calls every day, ranging from spam and scams to delivery people and financial service companies trying to contact them.

Lorenzo Franceschi-Bicchierai / TechCrunch: Oracle warns customers of a critical PeopleSoft flaw after ShinyHunters claimed breaches of 100+ organizations using PeopleSoft; Oracle has not issued a patch  —  Oracle warned its corporate customers that there is a critical-rated vulnerability in its PeopleSoft software …

Christina Kyriasoglou / Bloomberg: Infineon plans to open a €5B chip factory in Germany backed by EU subsidies, its largest single investment, on July 2 as Europe seeks to boost chip production  —  Infineon Technologies AG is preparing to open its largest single investment, a €5 billion ($5.8 billion) …

Avataar AI's distilled video model is priced at $0.005 for every second of generation

Equal AI said that its AI-powered call assistant now has over a million monthly active users.