← Tech Brew Ride Home
Tech Brew Ride Home
Xbox Decimated By Layoffs
00:22:20
Microsoft laid off ~4,800 employees and gutted Xbox by 3,200 jobs while divesting five studios. Samsung's profit rocketed past Nvidia's, Meta faced a $1.4 trillion lawsuit demand, xAI rebranded to SpaceXAI, and Anthropic researchers found a hidden "J-space" inside Claude.
Loading summary
Transcript1 lines
- [00:04]
A
Welcome to the tech brew right home for Tuesday, July 7, 2026. I'm Brian McCullough. Today, Microsoft is laying off thousands more employees and basically gutting Xbox. Samsung's profit rocketed past. Nvidia's Meta is facing a $1.4 trillion lawsuit demand Xai is rebranding to SpaceX. AI and anthropic researchers have found a hidden J space inside of Claude here's what you missed today in the world of tech. Every day, shareholders meet to discuss important matters about the companies you invest in. Now you can make your voice heard, too. Vanguard Investor Choice makes it easy to set your proxy voting preference for your eligible Vanguard index funds. Whether you hold a Vanguard fund directly or through another brokerage firm, all it takes is a few clicks to select your proxy voting preference and be heard on important shareholder topics like executive pay and Director elections. Visit vanguard.com investor choice to learn more. It's your shares. It's your voice. It's easy. Vanguard investors own shares of our index funds, and those funds own shares of the companies they invest in. Vanguard Marketing Corporation Distributor Microsoft is laying off around 4,800 employees, or around 2.1% of its workforce. Most of the affected employees are in sales or in Xbox. More on that later, quoting the Verge A year after cutting around 9,100 employees, Microsoft is making further layoffs today as it begins its new fiscal year. The software maker is laying off around 4,800 employees today, approximately 2.1% of its workforce. Most of the employees affected by today's cuts are in Microsoft's commercial sales business or the company's Xbox division. In an internal memo to employees, Amy Coleman, Microsoft's chief people officer, blamed the job losses on a changing technology industry and the need to adjust resources and roles to shift how we operate to respond to how AI is impacting companies like Microsoft. I also want to be direct that the roles eliminated today are not being replaced by AI, says Coleman. At the same time, what is true is that AI is changing how work gets done. Decisions like these are never easy, and you have my commitment that we are constantly looking for ways to reduce the need for job elimination, says Coleman. Whenever possible, our priority is to place people into new roles aligned to the company's highest priorities and greatest areas of opportunity. Over the past year, we have redeployed more than 4,000 employees into new roles and including another 500 this month. Microsoft had also been trying to avoid layoffs with its voluntary retirement program. U.S. employees whose combined years of service added to their age totals 70 or more were eligible for voluntary retirement, and the package will include five years of access to Microsoft's healthcare coverage, a lump sum cash severance payment and six months of vesting for unvested stock options. More than 30% of eligible employees chose to participate in our recent voluntary retirement program, and we will continue exploring similar approaches in the future, says Coleman. While this doesn't change the difficulty of today's news, we will continue to do everything we can to create opportunities for our people, reduce the need for job eliminations where possible, and responsibly support those affected with care and support. End quote. But as I said, things are even worse inside of Xbox, where 3,200 of those jobs will be eliminated over the next year, including 1600 just this week. Microsoft also plans to divest up to five studios, including Ninja Theory, quoting Bloomberg. Our business today is not healthy, xbox Chief Executive Officer Asha Sharma wrote in a note to staff Monday morning, adding that Xbox is operating at margins three to 10 times lower than comparable businesses. We must reset Xbox. The goal at Xbox is to streamline the business and reinvest in bigger projects, sharma wrote. It's the most significant move yet for Sharma, who became Xbox CEO in February, inheriting what she described as a business in distress. Despite making big investments, including the $69 billion purchase of Activision Blizzard in 2023, Xbox has struggled to release hit games, seen hardware sales decline significantly and face an increasingly tumultuous market. Last month, Sharma wrote in a memo to employees that Xbox's accountability margin, the metric Microsoft uses to reflect profit margin, had fallen to 3% and that annual revenue had plummeted. Going forward, this cannot continue, she wrote. Then the changes at Xbox will have significant repercussions throughout the video game industry. Last month, Sharma told staff that the division is expecting to pay five times as much for memory and storage in 2027 than it did in 2024. Xbox will sell the studio's Ninja Theory, maker of Hellblade, and Undead Labs, maker of State of Decay, to undisclosed buyers. Both studios will continue working on their current game projects, SUNA and State of Decay 3 with Xbox. Two other studios, Double Fine, maker of Psychonauts and Compulsion Games, maker of south of Midnight, will be spun out and returned to private ownership under their founders. Sharma wrote that these two studios will receive Runway funding and full ownership of the intellectual property they have developed, including their catalogs of previous games. All five of the studios were purchased by Xbox under former CEO Phil Spencer, who oversaw a spending spree over the last decade that scooped up both small game makers and large publishers such as Activision Blizzard and ZeniMax, the parent company of Arkane. His hope was to use their games to boost the popularity of Xbox Game Pass, the company's Netflix like subscription program for video games. But growth of the service plateaued. Sharma wrote that those acquisitions created meaningful value, but they did not grow at the pace we expected, adding that in a typical year, Xbox was losing 64 cents for every dollar it invested in the business. It is neither possible nor desirable to own every great independent studio, she wrote. As we reset Xbox, we will help independent creators succeed by providing open development tools and audiences to realize their vision. Sharma wrote that no publicly announced projects will be canceled as part of the restructuring, but the cuts will take place across the entire organization, including the studios that Xbox is retaining, in part to focus on higher priority projects. ZeniMax will face a significant overhaul and will pivot to focus on its biggest franchises Fallout, the Elder Scrolls, Doom, Quake and Wolfenstein. Samsung is estimating Q2 operating profit of around $58.44 billion, a 19 fold jump in profit from a year earlier and above, a $57.02 billion estimate. This would make Samsung the most profitable company in the world, overtaking Nvidia. Note that I said profitable. We're not talking revenue here. Nvidia is still making more money in terms of revenue, but Samsung has made more money in terms of gross operating profit, Quoting the Journal the strong results from the world's largest memory chip maker are likely to help ease recent market concerns about the sustainability of AI computing capacity. Spending shares in Samsung more than doubled from April to June, only to lose momentum this month as investors jitters resurfaced. Samsung has extended its run of record quarterly revenue and operating profit since the final quarter of 2025, its chip making division expected to fuel even stronger earnings growth this year, many analysts said. Citing booming demand for AI chips. The South Korean technology company said in a preliminary earnings report Tuesday that its operating profit would likely reach a record high of about $58.47 billion for the first three months ended June. That would be a 56% jump from the previous quarter's record. Some analysts said that Samsung's operating profit would have topped its own estimate had it not booked a provision for special bonuses payable early next year to employees in its chipmaking division. Samsung in May agreed to allocate 10.5% of its semiconductor division's annual operating profit to special bonuses contingent on the company meeting certain profitability benchmarks. Citigroup analyst Peter Lee said that Samsung's recent share price pullback may be a technical correction. The stock has surged more than 150% this year, supported by better than expected server DRAM chip pricing, driven by strong demand for AI central processing units. We believe memory fundamentals are intact and server DRAM pricing has been outperforming on strong CPU demand demand, li said, raising his 2026 operating profit forecast for Samsung to 401 trillion yuan from 334 trillion yuan. The company is scheduled to release full quarterly results, including a breakdown of earnings by business segment, later this month. End quote. Turns out that Elon Musk likes rebranding about as much as Google does because XAI is changing its name to SpaceX A and has unveiled a new logo. Elon said in May that XAI would be dissolved as a separate company and become the AI products company for SpaceX. Quoting Business Insider, the handle for the XAI account changed to SpaceXai on Monday. The account also shared a video of the XAI logo getting folded into a new SpaceX AI logo. While SpaceX is best known for its rockets and extraterrestrial ambitions, its IPO filings revealed just how much it was investing in AI. The company's capital expenditures on AI were $12.7 billion in 2025, or more than three times what it spent on its space and connectivity segments, which include Starlink, its satellite Internet service. Its AI segment has been a net loss for the company, but SpaceX believes it has the most potential, saying the total addressable market is the largest in human history. SpaceX said it plans to deploy AI compute satellites or data centers in space as early as 2028. The company has also landed some big AI infrastructure deals, with Anthropic agreeing to pay SpaceX $1.25 billion a month for access to compute power at its Colossus data centers, and Google agreeing to pay $920 million a month. End quote. Ever spent hours of your workday tracking down information only to find that it wasn't documented at all? It's an age old cor conundrum. Critical knowledge isn't available or accessible to the folks who need it. That's the very problem our sponsor Scribe was built to fix. Scribe is a workflow AI platform that automatically turns your workflows into clear cut documentation. All you do is turn on the Scribe browser extension or desktop app and go through your processes as you would normally. Scribe builds a guide as you go, capturing every click, step and screenshot automatically. So what would have taken hours gets done in under a minute. To see what Scribe could look like for your org, head to Scribe how ridehome and mention ride home for your first month of Scribe capture free on select plans. That's S C R I B E How ridehome. When your company deploys a customer facing AI that misses its mark, who do you think those customers blame? Well, According to the 2026 Delight AI Index, 83% of customers surveyed blamed the brand, not the tech they were using. That's why it's important to have tech you can trust. Delight AI is a customer facing AI concierge that delivers hyper personalized experiences on your behalf with zero touch improvement. It can continuously monitor its own performance, find failure patterns before they spread, write the fix and ship it. It gets smarter with every conversation automatically. The longer it runs, the more your customers can trust it. To learn more, just head to delight AIBrew. That's delight AIBrew In a court filing, Meta says four US states suing it over claims it designed Facebook and Instagram to addict youth and mislead the public are seeking $1.4 trillion in penalties. Meta's market cap is around 1.5 trillion. Quoting Reuters A sanction of that size has no analog in the history of consumer product enforcement, the company said in a filing. The state's filings are sealed, but at a court hearing in June, they said they were calculating the penalties by multiplying the number of violations by fine amounts set by state law. The number of violations is based on the estimated number of teens and young users affected by Meta's actions, the states said. 29 states have sued Meta in federal court, most of them alleging the company violated the federal Children's Online Privacy Protection act by collecting data from children without proper parental consent. The trial in August before U.S. district Judge Yvonne Gonzalez Rogers will address all claims brought under that law, plus the four states allegations that the company violated their state laws protecting consumers by misleading them about the safety of their platforms. Meta has denied the allegations, saying the attorneys general have no evidence it misled consumers about its platform's alleged addictiveness because social media addiction is not an established psychiatric condition and therefore statements that its platforms were not addictive could not be false. A further 14 states have brought claims under their own laws, which will be heard at a separate trial in February. Last month, Rogers rejected Meta's bid to cancel the trial, saying there remained factual disputes over whether its social media platforms were addictive, whether Meta falsely denied it designed them that way, and whether it partially directed the platforms at children. California Attorney General Rob Bonta said after Rogers, ruling that Meta was putting profits ahead of children's safety and breaking consumer protection laws, promising to hold the company fully accountable for its role in the teen mental health crisis. Meta, Snapchat and Parentsnap, YouTube and parent Alphabet and TikTok and parent ByteDance are facing thousands of lawsuits in both federal and state court over claims they knowingly design their platforms to have features that addict children and teens fueling a mental health crisis. States across the country have sued the companies, some as part of the case, before Rogers and others in their home state courts. New Mexico was the first to go to trial, and a jury awarded the state $375 million in March after finding the company had misled New Mexico consumers. Finally today, this deep into the LLM era of AI, it's somewhat striking to stop and remember how little we actually understand how this stuff works. Anthropic researchers have detailed what they are calling J space, a small set of neural patterns inside of Claude that they say reveals internal thoughts that don't appear in the model's output. Quoting VentureBeat Anthropic, the artificial intelligence company published a sweeping research paper on Sunday revealing that its Claude language models have spontaneously developed an internal structure that mirrors one of the most influential theories of how human consciousness works. The finding, which the company says has already begun reshaping how it monitors AI systems for safety risks, lands amid an intensifying scientific debate over whether machines can possess anything resembling a mind. The 16 author study, titled Verbalize Representations from a Global Workspace in Language Models, describes how anthropics researchers used a new mathematical technique to peer inside Claude's neural network and discovered what they call a J space, a small privileged zone of internal activity where the model holds concepts it can report on, reason with and direct at will, surrounded by a much larger ocean of automatic processing it cannot access or articulate. The researchers present evidence that an analogous functional distinction has emerged in modern AI models to what exists in humans, specifically observing that language models maintain a privileged set of internal representations available for report modulation and flexible internal reasoning atop a much larger volume of automatic processing. The parallel they draw is to global workspace theory, an influential account from neuroscience first proposed by cognitive scientist Bernard Bars. In the theory, the brain operates like a theater doz of specialized processors work in parallel backstage, but only a tiny spotlight of information at any moment gets broadcast to the whole theater, becoming what we experience as conscious thought. Anthropic says the J space achieves many of the same functional properties, even though the underlying Architecture of language models look nothing like a brain at the heart of the discovery is a new interpretability tool the researchers call the Jacobian lens, or J lens. The technique works by computing for each word in the model's vocabulary the average mathematical effect that a given internal activity pattern would have on making the model say that word at some point in the future. The crucial distinction is between what the model is saying and what is on its mind. When a JSpace pattern activates, it does not mean the model is about to say that word, just that the concept is available for the model to think with. Unlike a chain of thought scratch pad, the J space operates silently in the model's internal neural activations, allowing it to hold a concept without writing it down. Critically, the researchers report that this workspace was not deliberately engineered. It emerged on its own during Claude's training process, when the team applied the jlens across Claude's layers of computation. The model's processing divided into three distinct an early sensory zone, where raw input is parsed a middle workspace band where abstract, persistent concepts appear things like recognizing a face in an image, noticing a bug in code, or internally flagging search results as a prompt injection and a final motor zone, where internal representations collapse into whatever specific word the model is about to output. The researchers close the piece with a provocation that is likely to reverberate well beyond the interpretability community. That such a structure exists at all in language models is striking, they write. It suggests that the functional architecture associated with conscious access is not an accident of biological implementation, but a solution that learning systems converge on when faced with the right computational pressures. If the mind is an ocean, as the paper's authors write in their opening line, they have spent the last year charting its currents in a system that has no biology, no evolution, and no body, and found beneath the surface a structure that looks unsettlingly like the one we used to think. End quote. Nothing more for you today. Talk to you tomorrow.