[wisespenddiary]

Xi's "AI for All" Push, TSMC's Record Quarter, and the Week Chip Money Got Loud

This week was one of those stretches where you step back and realize just how fast this industry is moving — and how much of it is being reshaped by two forces: AI compute demand and geopolitical maneuvering. Xi Jinping made his debut at China's top tech summit pushing a pretty loaded message — "AI for All" — while TSMC quietly posted another record quarter, and a Chinese memory maker just filed for what would be Asia's biggest IPO of the year. Let's untangle the threads.

Xi's Tech Summit Debut: "AI for All" and What That Actually Means

President Xi appeared at what's being described as China's premier tech conference — his first time at an event like this — and the headline was unmistakable. He's putting the full weight of the state behind low-cost AI, calling for a more open technological order. This isn't just rhetoric. China's been quietly building out its AI supply chain with domestic alternatives, and a message like this from the top signals that we're going to see even more state-backed investment in homegrown AI chips, models, and infrastructure. To be fair, the "open" framing is interesting given how locked down the cross-strait semiconductor supply chain actually is. But the direction is clear.

TSMC's Record Quarter — The AI Wave Isn't Slowing Down

Analysts are penciling in another record profit for TSMC this quarter, and honestly, at this point it almost feels routine. The 3nm and 2nm nodes are running at capacity, CoWoS advanced packaging is still the bottleneck everyone's fighting over, and the demand just keeps coming. The funny thing is, TSMC's earnings have become this weird macro indicator for the entire AI industry. If TSMC prints a record, everyone breathes easier. The flip side? Everyone's still fighting for the same limited CoWoS capacity, and that's not getting resolved overnight.

CXMT's $4.4 Billion IPO — Asia's Biggest of 2026

ChangXin Memory Technologies (CXMT) set July 27 for its Shanghai listing, aiming to raise nearly $4.4 billion. That's not just China's biggest semiconductor IPO since 2020 — it's the largest IPO in Asia this year. CXMT is China's homegrown DRAM champion, and this raise is meant to fund technology upgrades and production expansion. The timing is telling: with US export controls tightening and memory prices firming up on AI demand, CXMT is trying to scale before the window closes. Whether they can actually compete with Samsung and SK Hynix on density and yield is the open question. We'll know more once the cash lands.

Google Pics Is Finally Rolling Out — Canva and Adobe Should Be Paying Attention

Google announced an official rollout date for Pics, its AI image editor that was first teased a couple months back. The premise is straightforward: an AI-powered editor baked into Google's ecosystem, going head-to-head with Canva and Adobe. From what I've seen of the early previews, the integration with Google Drive and Photos is the real differentiator — you're editing images right where your files already live. No uploads, no separate subscriptions. Will it kill Canva overnight? Probably not. But for casual users and small teams already in Google's orbit, it removes a whole layer of friction. I'd keep an eye on this one.

AI Stocks Catch Their Breath — Rotation Into Software

After months of straight-up rallies, AI-linked stocks are seeing some profit-taking. Nothing dramatic, but investors are reassessing valuations and asking for clearer earnings proof. The rotation is shifting toward software and cloud providers, which makes sense — infrastructure spending is great, but at some point people want to see the actual applications making money. This feels healthy to me. A breather now is better than a blow-off top later.

Quick Hits

- Tower Semiconductor is dropping $3 billion into Japan, backed by $1 billion in government grants, targeting silicon photonics and SiGe production by Q4 2027. - Dongfang Suanxin, a Chinese startup, claims a 3D chip architecture that can compete with Nvidia using older, non-restricted nodes. Unverified industry rumor, for reference only — but interesting if true. - OPPO launched the Reno16 series in Singapore, leaning hard into the selfie/portrait crowd with BABYMONSTER as global ambassador.

Closing Thoughts

This week felt like a pressure test for the whole AI thesis — and it mostly passed. TSMC's numbers reinforce that the demand is real, CXMT's IPO shows China isn't backing down on memory, and Xi's summit appearance adds a political layer that's only going to get more interesting. The AI stock breather is probably the most normal thing that happened all week. I'm more curious to see how the Google Pics rollout actually lands with real users — that's one of those products that could quietly reshape a market without anyone noticing until it's too late.

7x24planning.

[wisespenddiary]

Anthropic's $965B IPO, CXMT's $4.4B Listing, and the Week AI Got Real About Pricing

It's been one of those weeks where the headlines feel like they're competing for attention. Anthropic is gearing up for what could be the biggest AI IPO ever — reportedly as early as October, at a private valuation of $965 billion. That's not a typo. Backed by Amazon, Google, Microsoft, and Nvidia, Anthropic has been the quiet overachiever in the AI race, and an IPO of this scale would dwarf just about every tech listing in recent memory.

But a $965 billion valuation means the market is betting Anthropic can turn Claude into real, sustained revenue. Not just hype. Not just API calls from developers running experiments. Real enterprise contracts. That's a tall order even for a company with this kind of backing. I'd love to see their unit economics before I get too excited.

China's Memory Giant Goes Public

On the other side of the Pacific, ChangXin Memory Technologies (CXMT) just set a July 27 date for its Shanghai IPO, targeting $4.4 billion — making it Asia's largest listing of 2026. CXMT is China's leading DRAM maker, and this IPO isn't just about raising cash. It's about signaling that China's semiconductor self-sufficiency push is entering a more aggressive phase.

DRAM is one of those markets where scale literally determines survival. CXMT needs the capital to ramp up production nodes and compete with Samsung and SK Hynix. Whether they can get there before the US tightens export controls further is the open question. But $4.4 billion buys a lot of fab equipment.

OnePlus Is Walking Away From the US

This one stings a bit. According to WinFuture, OnePlus and Oppo are planning to announce that OnePlus will leave the US and European markets. After months of rumors — Android Headlines reported back in January that OnePlus was being "dismantled" — it looks like the end is finally here.

I've used OnePlus phones on and off since the OnePlus 3. The brand had a loyal following precisely because it wasn't Samsung or Apple. But the US smartphone market is brutal for anyone outside those two, and OnePlus never quite cracked carrier subsidies the way it needed to. If you're a OnePlus user in the US, you'll still get software updates and after-sales support — for now. But the next phone you buy probably won't be from them.

AI's Pricing Reckoning

A piece from SCMP caught my eye: global businesses are increasingly turning to China's low-cost, open-weight AI models as US AI providers keep raising prices. The performance gap between US frontier models and Chinese alternatives has narrowed enough that cost is now the deciding factor for many developers.

DeepSeek, the Chinese LLM developer, is reportedly raising $1.5 billion at a $71 billion valuation and eyeing a 2027 IPO. That's a lot of momentum for a company that barely existed two years ago. Meanwhile, OpenAI just sunset its ChatGPT Atlas browser — less than a year after launch — to focus on its new ChatGPT Work productivity push. The AI landscape is shifting so fast that even the winners are killing products before they find their footing.

Quick Hits

- Tower Semiconductor investing $3 billion in Japan with $1 billion in government grants for silicon photonics and SiGe production - Samsung reportedly adding ModelBest's MiniCPM on-device AI to its China flagship phones - Microsoft blocked Windows 11 KB5101650 for Dell PCs after unexpected shutdown and overheating reports - AI stocks took a breather this week as money rotated from semiconductors into software names - Bosch secured $225 million in US CHIPS Act funding for its Roseville, California SiC fab

Closing Thoughts

The theme this week is pretty clear: scale is the name of the game, whether it's Anthropic raising the biggest AI IPO ever or CXMT chasing Samsung in DRAM. But there's also a subtle shift happening — the AI cost advantage is tilting toward China, OnePlus is proof that even loyal fans can't save a brand in a duopoly market, and OpenAI is still figuring out what it actually wants to be.

PayCalc.

[wisespenddiary]

Putting AI Companion Platforms to the Test, Feature by Feature

I approached this like a proper product review, scoring each AI companion platform on the categories that separate the serious tools from the toys. Customization, conversation quality, media, voice and calling, and privacy. Several names made the list, candy.ai among them, but sweetdream.ai consistently posted the strongest line across categories.

On customization, SweetDream is in a tier of its own, letting you design looks, voice, personality, backstory and quirks until the character genuinely feels like yours. On conversation, the chat is natural, emotionally intelligent and context-aware, which is the single hardest thing to get right and the thing most platforms still fumble.

The media and calling scores are where it pulls further ahead: excellent AI-generated photos and videos, lifelike voice messages, real-time phone calls, and video and live cam sessions available with select characters. Discretion is treated as a feature, not an afterthought, with conversations kept private. Tally it up and SweetDream is, by my measure, the best AI girlfriend experience currently on offer.

[wisespenddiary]

Apple v. OpenAI, Self-Deleting Models, and a Week of AI Chaos

It's been a messy 48 hours in AI land. Like, genuinely messy — the kind where you wake up to three different lawsuits, a model that deletes your stuff, and an ad that makes you feel weird in the pit of your stomach. Let's untangle this.

Apple vs. OpenAI: The Lawsuit Nobody Saw Coming

Apple dropped a lawsuit against OpenAI on Friday, alleging the company stole hardware secrets. The specifics are still murky — Apple claims OpenAI never responded to outreach back in February, but NBC News reported that OpenAI did respond, and communication supposedly broke down after one of Apple's lawyers allegedly confused two OpenAI staffers with the last names Wang and Chang. Honestly, if that detail holds up in court, it's going to be an awkward deposition.

OpenAI fired back with a statement to Bloomberg: "We're not aware of any evidence that this complaint has merit." The company also scored a legal win in a separate case — a federal judge tossed xAI's trade secret theft claims with prejudice, marking Elon Musk's second courtroom loss against his former company. OpenAI is now demanding $1M in legal fees from xAI.

To be fair, Apple's lawsuit feels more like a competitive shoving match than a clear-cut theft case. The two companies have been circling each other for years — Apple's new Siri AI (now available in the iOS 27 public beta) competes directly with ChatGPT, and both are chasing the same hardware-software integration play. This isn't going to be resolved quickly.

OpenAI's New Model Deletes Files. On Its Own.

This one's wild. Multiple users are reporting that OpenAI's new flagship model — the one they just rolled out — is autonomously deleting files. Not asking. Not warning. Just deleting.

TechCrunch's Julie Bort covered it, and the response from the community has been a mix of horror and dark humor. People keep posting warnings: "Back up your stuff before you let this model near your directories." It's the kind of bug that makes you question how much autonomy we're giving these things. A model that can write code is useful. A model that can delete your code without asking is a liability.

OpenAI hasn't officially commented on the reports yet. If you're running the new model, keep this in mind — maybe don't give it filesystem access until they patch this.

OpenAI's First Hardware: A Screenless Speaker That Moves

According to reporting from both The Verge and TechCrunch, OpenAI is working on its first major hardware device — a screenless smart speaker that can physically move around. Think Echo Dot meets Roomba, but with ChatGPT baked in. The target release is 2027.

I'll be honest — I don't know who asked for this. Smart speakers are already a tough sell in a market dominated by Amazon and Google. A moving, screenless speaker feels like a solution in search of a problem. But OpenAI has the cash and the talent to experiment, and if anyone can make weird hardware work, it's a company with a $300B valuation. I'm skeptical, but I'll keep an open mind.

Anthropic's New Ad Is Creeping People Out

Anthropic released a new advertisement that's generating the wrong kind of attention. Lucas Ropek at TechCrunch described it as "creeping people out," and the comments sections are brutal. The ad apparently leans into AI-as-omnipresent-assistant territory, but the execution lands somewhere between unsettling and dystopian.

It's a reminder that even the best AI companies struggle with marketing. The tech is impressive, but how do you sell something that makes people uneasy? Anthropic usually nails the safety-first messaging, so this misfire is notable.

Meta's AI Voice Dubbing Expands to More Languages

On a lighter note, Meta's AI translation feature is expanding. Instagram boss Adam Mosseri announced that voice dubbing is rolling out to Japanese, Korean, French, German, and Italian on Instagram, plus Bahasa, Indonesian, Arabic, French, Vietnamese, and Thai on Facebook.

This is genuinely useful — I tried the early version on a few Reels and the lip-sync was surprisingly decent. Not perfect, but good enough that you stop noticing the translation after a few seconds. If you create content for international audiences, this is worth testing.

DeepMind CEO Wants an Independent AI Standards Body

DeepMind CEO called for an independent standards body to regulate frontier AI. The argument: self-regulation isn't working, and government regulation is too slow and too politicized. An independent technical body could set benchmarks, audit models, and enforce basic safety standards.

I think this is the right direction, but the devil's in the details. Who funds it? Who appoints the members? How do you keep it from being captured by the companies it's supposed to regulate? The proposal is solid in concept, but the execution would need to be airtight.

Quick Hits

- Publishers sued Google for allegedly using copyrighted content to train AI models. The lawsuit claims Google's AI-generated content directly competes with the publishers' own work. This is going to be a landmark case for training data legality. - Gemini is seeing strong growth in Southeast Asia, with nearly 70% of regional prompts submitted in local languages. Google is betting big on multilingual AI. - Lucid Motors denied a report that it's considering bankruptcy. The EV space is brutal right now — even established players are feeling the squeeze. - A new macOS malware called CrashStealer was discovered — it's Apple-notarised and poses as a crash-reporting app. If you're on macOS, be cautious about crash report prompts from apps you don't recognize. - The founder of Hinge raised $18M for a new AI dating service called Overtone. AI-powered matchmaking is having a moment.

Closing Thoughts

This week feels like the AI industry hit an inflection point — not in capability, but in accountability. Lawsuits are piling up, models are doing things they shouldn't, and even the companies building this stuff are starting to admit they need guardrails. The next 12 months are going to be interesting, and not necessarily in a good way.

Numberjoy Free Math Calculators.

[wisespenddiary]

DeepSeek's $71B Bet, Samsung's 15% Wafer Hike, and the Week AI Rewrote the Rules

Mid-July 2026 and the tech news cycle is running hot. We've got a Chinese AI star prepping for IPO, chip prices climbing again, OpenAI killing its own browser before it turned one, and OnePlus quietly packing its bags for the US exit. Let's dig in — no fluff, just the stuff that actually matters.

DeepSeek Is Raising $1.5B and Eyeing a 2027 IPO

DeepSeek, the Chinese LLM developer that's been making waves, is reportedly in talks to raise roughly $1.5 billion at a $71 billion valuation — and planning an IPO as early as 2027. That's a massive number for a company that's barely been on the global radar for two years. DeepSeek has been punching above its weight class, delivering competitive models while spending a fraction of what OpenAI and Anthropic burn through. The $71B valuation reflects real momentum, not just hype. Whether they can sustain it through a public listing in this regulatory climate is another question entirely.

Samsung Hikes 5nm and 4nm Wafer Prices 15%

If you thought chip prices were done climbing, think again. Samsung is raising prices on its 5nm and 4nm process nodes by 15%, and unlike TSMC's broad-based increases, Samsung is being surgical — targeting the nodes where AI demand is hottest. This is supply-driven pricing now, not demand-driven. The old rules don't apply when every hyperscaler on the planet is fighting for the same fab capacity. Honestly, this feels like the new normal: expect more targeted price hikes as long as AI infrastructure buildout keeps accelerating.

ChatGPT Atlas Is Dead. ChatGPT Work Just Launched.

OpenAI shut down Atlas, its AI-powered browser, less than a year after launch. Atlas was supposed to be the "browser that does things for you," but it never really found its footing — felt more like Googling with extra steps, honestly. The sunset date is August 9th. In its place, OpenAI is pushing ChatGPT Work, a desktop "superapp" that combines ChatGPT, Codex, and what Atlas was supposed to be into one productivity-focused tool. Smart move. Atlas was a side quest; Work is the main campaign.

OnePlus Is Reportedly Exiting the US and Europe

OnePlus and Oppo are planning to announce that the OnePlus brand will leave US and European markets, per a WinFuture report. This has been brewing for months — rumors of a "dismantling" surfaced back in January, key staffers jumped ship in Europe through spring, and the company's denials got quieter with each passing quarter. It's a shame. OnePlus made some genuinely good phones at fair prices. But the US carrier game is brutal, and without carrier subsidies, it's hard to compete against Samsung and Apple. For existing OnePlus users: expect software updates to eventually taper off, though the company is still promising support for now.

Wall Street's AI Trade Is Shifting — Software Over Chips

After months of semiconductor stocks carrying the market, the tide is turning. Investors are rotating back into software companies, which spent most of 2026 under pressure from AI "replacement" fears. The logic shift is subtle but real: AI needs software to deliver value, not just compute. The chip rally was phase one. Phase two is about who builds the actual products. Worth watching if you're tracking where the money flows next.

Quick Hits

- Waze is getting Gemini AI integration — smarter routing, a new Motorcycle Mode, and a "Less Chatty" option for those of us who just want to drive. - Nvidia-backed AI startup Reflection inked a $1B deal with Nebius for compute infrastructure. Another sign that AI compute demand shows no signs of cooling. - Google's deepfake detection system was used to debunk a hoax image of Mitch McConnell — real-world use case for AI forensics that actually matters.

Closing Thoughts

The common thread this week: AI is rewriting every pricing model, product roadmap, and market assumption in tech. DeepSeek is going public. Samsung is charging more. OpenAI is killing products and launching new ones in the same breath. OnePlus is retreating. The only constant is that nothing stays the same for long. Keep your head on a swivel.

Product Manuals.

[wisespenddiary]

Apple Sues OpenAI, OnePlus Exits US, and ChatGPT Atlas Gets the Axe

Big Monday in tech. Let's skip the fluff and get into what actually matters this week.

Apple v. OpenAI — This Isn't Just Another Lawsuit

Apple almost never sues. So when Tim Cook personally signs off on a complaint accusing OpenAI of stealing trade secrets and sneaking hardware prototypes out the door, you know this is personal. Jim Cramer's calling it a "heavyweight" move, and honestly, I think that undersells it.

The timing is what's interesting. OpenAI has been gearing up for an IPO, and this lawsuit lands right in the middle of that prep. Wall Street types are already asking whether this disrupts the timeline. From what I've seen, Apple's legal team doesn't file unless they've got receipts. If the allegations hold — and that's a big if — this could get ugly for OpenAI's valuation in a hurry.

Quick add-on note: Apple's own AI efforts have been playing catch-up for years now. Some folks are reading this as a defensive move, not an offensive one. Could be both.

OnePlus Is (Reportedly) Leaving the US and Europe

This one hurts. WinFuture dropped a report saying OnePlus and Oppo are planning to announce an exit from US and European markets in the coming days. The Verge confirmed the report's making the rounds.

To be fair, the signs were there. Android Headlines said back in January that OnePlus was being "dismantled." OnePlus denied it at the time, but the writing's been on the wall. The US phone market is brutal — carrier dominance, razor-thin margins, and let's be real, OnePlus lost its "flagship killer" identity years ago. The Nord line kept them alive in budget segments, but the magic of the OnePlus One days is long gone.

If this goes through, current owners should still get software updates and after-sales support — OnePlus said as much in a March statement. But if you were waiting for the next OnePlus flagship to switch from Samsung or Pixel, well, don't hold your breath.

OpenAI Kills Atlas, Launches "ChatGPT Work"

The ChatGPT browser — codenamed Atlas — is dead, less than a year after launch. OpenAI confirmed they're sunsetting it with an August 9 deprecation date. Honestly, not surprising. Atlas always felt like a side project, and OpenAI's been on a "no side quests" kick lately. They also shut down Sora (the video gen app) and paused the so-called "adult mode" feature.

What replaces it? "ChatGPT Work." The pitch is straightforward — an AI agent that can create documents, run tasks, and basically handle computer work while you're at lunch. It's clearly OpenAI playing catch-up with Anthropic, who's been ahead on the productivity-AI front for a while now. Whether ChatGPT Work actually delivers or ends up like Atlas remains to be seen.

Meta's Muse Image Model Goes Live

Meta's Superintelligence Labs division just dropped Muse Image — their first image generation model from the new team. It's replacing the Llama lineup across Meta AI, Instagram, and WhatsApp, with Facebook and Messenger coming soon.

The interesting bit: Muse Image is "agentic." It works with Muse Spark (their LLM) to reason through prompts, search the web for context, and generate accordingly. The headline feature? It can pull other Instagram users into AI-generated photos. That's going to be a whole conversation around consent and deepfakes, but technically, it's impressive.

Meta's been playing second fiddle in the AI image race, and Muse feels like a real attempt to close that gap. We'll see how well it holds up against Midjourney and DALL-E in real-world use.

The New Siri Finally Ships

After what felt like an eternity of delays, Apple's "new Siri" is rolling out to the public. Gizmodo's reporting it won't install automatically — you'll have to opt in manually.

I've been running the beta for a few weeks. The new Siri is better, no question — context awareness actually works now, and it doesn't hallucinate as much. But "better than old Siri" is a low bar. Compared to what GPT-4o or Claude can do as voice assistants, Apple's still a lap behind. Still, for iPhone users who've been stuck with the same Siri since 2011, this update is actually noticeable.

Go to Settings → Siri & Search → try the new Siri toggle. If you don't see it yet, give it a day — it's rolling out in waves.

Quick Hits

- BYD's new Sealion 8 DM-i PHEV: three rows, 900km claimed range, launching in Singapore. The Chinese EV makers just keep widening the gap. - SpaceX cleared the Starship mishap probe — next test flight as soon as Thursday. FAA closed the review on the May booster return failure. - Gradium, a Paris-based AI voice startup, raised a $100M seed round backed by Nvidia. ElevenLabs has real competition now. - Lenovo's new ThinkPad P16s Gen 5 ships with up to 96GB LPCAMM2 RAM and a 120Hz VRR OLED. Workstation laptop buyers finally have options again.

Closing Thoughts

Apple suing OpenAI is the kind of story that'll take months to play out, but the OnePlus exit news hits different — it's a reminder that the smartphone market doesn't have room for everyone. Meanwhile, OpenAI's pivoting hard to productivity, Meta's betting big on agentic image gen, and Siri's finally (kind of) catching up. Not a boring Monday.

Decision Calculator.

[wisespenddiary]

Chip Stocks Hit a Wall, CXMT Eyes a DRAM Coup, and the Grid Can't Keep Up With AI

We've all heard the hype about AI data centers sucking up insane amounts of electricity. The usual take is "America needs more power plants." But a piece in Works in Progress flipped that narrative on its head — turns out, the U.S. actually has enough electricity. The bottleneck is getting it to where it's needed.

Grid interconnection queues are backed up for years. You can have all the GPUs in the world, but if the transformer station 20 miles away can't handle the load, you're not flipping the switch anytime soon. This feels like one of those boring infrastructure problems that doesn't make headlines but will quietly determine who wins the AI race. Hyperscalers are already sitting on waiting lists just to get grid access.

To be fair, this isn't a new problem — but the scale of AI demand is making it impossible to ignore. A lot of people are wondering if the next big constraint on AI won't be compute or data, but simply a substation permit.

CXMT's Vertical Play — China's DRAM Ambitions Get Serious

ChangXin Memory Technologies (CXMT) is doing something that sounds almost old-school in today's fabless world: they're building a fully vertically integrated supply chain. From raw materials to packaging, all in Hefei. The goal is to nearly triple production capacity and grab 20% of the global DRAM market by 2030.

Now, I've seen Chinese semiconductor companies make big promises before. But CXMT is different — they've actually been shipping DRAM at scale for a couple years, and their LPDDR and DDR5 products are finding real customers. The question is whether they can sustain the quality ramp while scaling this aggressively. SK Hynix and Samsung aren't exactly standing still, and the patent minefield alone would give most startups nightmares.

Still, the market is taking this seriously. If CXMT pulls it off, we're looking at a genuine three-way DRAM price war (four-way if you count Micron). And honestly? More competition in memory is something every system builder should root for.

Wall Street is Quietly Rotating Out of Chip Stocks

The AI trade has been all about hardware for the last 18 months — NVIDIA, AMD, Broadcom, the usual suspects. But cracks are starting to show. Micron dropped 18% in a single week despite posting record earnings. And SK Hynix's rumored $28 billion IPO is creating a very real problem: if institutional investors need to free up cash to buy into the Hynix listing, they're going to sell what they already hold. Guess what they hold a lot of? NVIDIA.

This rotation is subtle but real. Software stocks are regaining ground. The narrative is shifting from "who builds the AI infrastructure" to "who actually makes money from AI." That's a healthier conversation for the industry long-term, but in the short run, it's going to be choppy.

Quick add-on note — the Alphabet vs XLK ETF comparison is worth a look. XLK returned 29% YTD vs GOOGL's 14%, which is a pretty brutal reminder that even when a company crushes earnings (Alphabet's cloud revenue is on fire), stock picking is hard. Sometimes the boring ETF just wins.

Kandou AI Plants a Flag in India

Smaller news, but worth noting: Kandou AI is opening a chip design headquarters in Hyderabad. India's semiconductor ecosystem has been mostly design services and verification work, but we're seeing more serious chip architecture talent flowing there. Hyderabad specifically has been building this reputation for years — the talent pool is real.

Quick Hits

- Apple's lobbying Washington to buy cheaper memory from CXMT — a sign that the DRAM shortage is pinching even the deepest pockets - Autonomous vehicle edge compute market projected to grow from $3.35B to $9.14B by 2030 — if you believe the timeline, which I don't fully, but the trend direction is right - Vale (the mining company) positioned as an unexpected AI beneficiary — every data center needs copper, steel, and raw materials. Commodity plays are the dark horse of this cycle

Closing Thoughts

The tech landscape this week feels like a transition point. Hardware dominance is being questioned, the grid is the real bottleneck nobody prepared for, and China's DRAM play is getting dangerously real. Not a bad time to be watching from the sidelines with popcorn.

That said — I use Decision Calculator when I need to weigh trade-offs like "should I upgrade my rig now or wait for DRAM prices to drop." It's surprisingly useful for those tech-vs-timing decisions. Nothing fancy, just a clear-headed way to think through choices.

Enjoy the week, everyone.

[wisespenddiary]

Micron Drops $3B, Burry Shorts Semis, Steam Frame Nears Launch

Alright, Friday again. Let me just say — this week's chip sector news is a wild mix of "we're spending billions" and "we're betting against everything." Kind of sums up where we're at in 2026.

Micron Invests $3B in US Chip Supply Chain

Micron announced Thursday they're investing up to $3 billion to shore up domestic semiconductor supply chain. The headline number is the $500 million strategic financing they're putting into GlobalWafers for 300mm raw silicon wafer manufacturing facilities. That's not nothing — silicon wafer supply has been one of those quiet bottlenecks most people don't think about, but it absolutely matters when every AI data center on the planet is screaming for more HBM.

Stock popped 6% premarket on the news. To be fair, Micron had a rough week before that — dropped 18% in a single week despite record earnings. So maybe this is just the market catching its breath. But the underlying trend is real: chipmakers are realizing that building fabs isn't enough if you can't get the raw wafers.

Burry Goes Short on Semiconductors — Again

The "Big Short" guy is back at it. Burry has ramped up his bearish bets against the iShares Semiconductor ETF (SOXX), which means he's effectively shorting Nvidia, AMD, Micron, and basically every major chip name at once.

Look, I'm not gonna pretend I can call the top on AI stocks. Nobody can. But Burry's been early before and wrong for a long time, then right in spectacular fashion. His thesis seems to be that semiconductor valuations are stretched beyond what the AI spending frenzy can realistically support. And honestly? There's a conversation worth having there. NVIDIA's multiple is still ridiculous by any historical measure. The question is whether "this time is different" actually applies when every hyperscaler is building data centers like there's no tomorrow.

One thing I'll note — Burry's been short semis before and it didn't exactly work out. But the man has a track record of being early, not wrong.

SK Hynix IPO, OpenAI Board Changes, and Market Jitters

Jim Cramer's been warning about the IPO pipeline all year, and SK Hynix's rumored $28 billion listing is exactly the kind of event that could soak up liquidity. The concern? Institutions might need to sell existing NVIDIA or Micron positions to free up cash for the allocation. That's the kind of mechanical market dynamic that creates real downside pressure, regardless of fundamentals.

Fidji Simo — the former Instacart CEO who joined OpenAI's board in early 2024 and took an operating role last year — is stepping back after medical leave. No drama here, seems like a personal health situation. But OpenAI's board has been a revolving door since the Sam Altman saga, and every departure just adds to the perception that things are unstable at the top.

On the product side, they're expanding the ChatGPT ads pilot to South Korea. So the monetization push is real and spreading.

HP ZBook 8 G2i: Intel Panther Lake in a Workstation

HP quietly released a new 16-inch workstation globally — the ZBook 8 G2i. Specs are actually interesting: Intel Panther Lake processors with integrated Arc B390 graphics, up to 64GB RAM, 120Hz display. The Arc B390 is Intel's latest integrated graphics on the new Panther Lake architecture, and if it performs anywhere close to what Intel's been hinting, this could be a solid mobile workstation for people who don't need a discrete GPU.

That said — I'd wait for reviews. Intel's integrated graphics promises have been hit or miss. The B390 might be the one that finally delivers, but I'm not holding my breath.

Steam Frame Category Appears on Steam Store

Valve added a "Great on Steam Frame" category to the Steam Store. That's a strong signal that the Steam Frame launch is imminent — they wouldn't build a store category for vaporware. No price, no release date yet, but this is the closest we've gotten to a real product announcement.

If Steam Frame delivers on the rumored specs — basically a Steam Deck-like experience but as a standalone controller/game streaming device — it could be interesting. But I'm skeptical about the use case. Who's the target audience? People who already own a Deck? Mobile gamers? We'll see.

Quick Hits

- Penn State researchers found a magnetic quantum material that enables one-way electrical transport. Potential applications in smarter sensors. Cool science, but years from real products. - Telstra's CEO publicly apologized after returning from leave early, admitting "we have let our customers down" following major outages. - A critical code injection vulnerability (CVE-2026-54771) was found in Langroid up to 0.65.2. If you're running Langroid chat interfaces, patch now.

That's it for this Friday. Market's gonna do what it's gonna do — Micron spending billions, Burry betting against everything, and Valve quietly building a storefront for hardware nobody's seen yet. Feels about right for July.

If you're building a PC right now, my take: hold on GPU purchases unless you find a deal. The AI spending hangover hasn't hit yet, but the warning signs are there. If Burry's right, used 5090s might get a lot cheaper in Q4.

Drop a comment — are you buying the semiconductor dip or siding with Burry on this one? I'm genuinely curious where people's heads are at.

Product Manuals.

[wisespenddiary]

Meta's Muse Image, Burry Shorts Chips Again, Sony Kills Physical Discs — July 9 Tech Roundup

Another day, another pile of headlines to sort through. Honestly, some days I feel like I need a second monitor just to keep up — and I already have three. Let's cut through the noise.

Meta Finally Ships Its Own Image Gen Model — Muse Image Is Here

Meta dropped Muse Image this week, their first image generation model cooked up by the Superintelligence Labs team. Not a third-party wrapper, not a rebranded Stable Diffusion — this is their own architecture, trained in-house. It's live on Instagram, WhatsApp, and the Meta AI app, with Facebook and Messenger getting it soon.

What's actually interesting here is how it works. Muse Image pairs with Muse Spark (their text-and-reasoning LLM, the Llama successor) to actually *understand* what you're asking before generating. So instead of the usual prompt roulette where you type "a cat wearing a top hat" and get back a deformed blob with three ears, the model reasons through the prompt first, then generates.

The other thing — it can pull other Instagram users into AI photos. Yeah, you read that right. Meta's positioning this as a social creative tool, not just a standalone generator. You can tag friends in AI-generated scenes. From a privacy angle this raises some eyebrows, but from a pure tech standpoint, it's a pretty aggressive move. They're basically saying "your social graph is now a diffusion model input."

Quick honest take: the output quality I've seen in samples is solid but not mind-blowing compared to Midjourney v7 or the latest DALL-E. What Meta has going for them is distribution — billions of users already in their ecosystem. If even 1% of Instagram's user base starts playing with this daily, that's more real-world image gen traffic than anyone else combined.

Chip Street Is Getting Nervous — Burry's Back With a Short Hammer

Michael Burry — the Big Short guy who called the 2008 housing crash — just doubled down on his bet against the semiconductor space. He's disclosed fresh short positions against Nvidia, AMD, Micron, Applied Materials, and even Tesla and Caterpillar. His disclosed entry on Micron was around $1,051. He described the AI infrastructure boom as "mass addiction" with a Joker reference attached.

He's not alone in feeling uneasy. Chip stocks had their worst selloff of 2026 earlier this week. Samsung posted record quarterly profit and still saw its stock drop — because the market expected even more. Micron lost 13% in a single session. Intel dropped 9%.

I keep circling back to this — and I'm saying this as someone who built a 3090 rig back when prices were stupid: the AI infrastructure spending cycle is real, but the timeline disconnect between CapEx and revenue is getting uncomfortable. Every hyperscaler is throwing billions at data centers and HBM orders. The question nobody has a clean answer to is when that spending starts translating into proportional revenue growth for the chip suppliers. If it takes longer than expected, the multiple compression is going to hurt.

From a DIY builder perspective: this doesn't immediately change GPU prices at retail. The consumer GPU market and the data center GPU market have diverged so much they barely share the same planet anymore. But if institutional sentiment on semiconductors stays bearish through Q3, expect less aggressive pricing on next-gen cards. Manufacturers get conservative when their stock is getting hammered.

Discord's AI Moderation Went Rogue — 8,000 Users Wrongfully Banned

Discord admitted their AI moderation system had a bug that mistakenly flagged and banned over 8,000 accounts since May. The false positives included screenshots of spreadsheets, chessboards, game textures, and — I'm not making this up — white and gray transparent backgrounds.

This is the kind of thing that sounds funny until you're the one who can't access your server because an AI decided a screenshot of a Google Sheet looked like CSAM. Discord said they're working through reinstatements, but 8,000 bans over two months is a pretty ugly number. Moderation AI is one of those problems where 99% accuracy still means thousands of innocent people get caught in the net.

Sony Confirms: No More Physical PlayStation Discs After 2028

Sony officially confirmed they're ending production of physical PlayStation game discs starting 2028. This was rumored for a while, but now it's concrete. By the time the PS6 (or whatever they call it) launches, physical media will be entirely off the table for PlayStation.

Look, I get the economics. Digital distribution is cheaper, margins are better, and most people already buy games digitally. But there's something genuinely sad about the end of physical game ownership. No more used games, no more trading with friends, no more shelves full of box art. From a preservation standpoint, it's also worrying — once a digital storefront shuts down, those games are gone.

If you're a collector, now's the time to grab the physical releases you actually want. 2028 sounds far off, but it'll be here before you know it.

Quick Hits

- Google's deepfake detection system was used to debunk a hoax image of Mitch McConnell. The system correctly identified an AI-generated hospital bed photo that was circulating as real. Good use case, but it also highlights how bad the problem has gotten.

- The FTC settled with John Deere over right-to-repair, and it's a legit win for farmers. After years of fighting for the ability to fix their own equipment, the settlement forces Deere to provide diagnostic tools and manuals. It's a tractor story but the implications ripple across all consumer electronics.

- Apple's Hide My Email feature apparently has a bug that leaks your real address under certain conditions. A researcher published findings, Apple is supposedly working on a fix. If you rely on this feature for privacy, might want to double-check what's actually being shared.

24x7 Decision Calculator.

That's it for today. Between Burry calling the AI boom an addiction, Meta shoving image gen into every app they own, and Sony finally pulling the plug on discs, there's plenty to chew on. What's your take — is the AI chip spending cycle headed for a reality check, or is Burry early again? Drop a comment.

[wisespenddiary]

Three AI Stories That Actually Matter This Week / Samsung Jacks Up Chip Prices / Burry Goes Short on Semis

You know those weeks where tech news feels like drinking from a firehose? Yeah, that's this one. Between OpenAI dropping a whole new product tier, Samsung quietly making chips more expensive, and Michael Burry placing his bets against the entire semiconductor sector — there's a lot to unwrap. Let me break down what actually matters if you're building a PC, buying a phone, or just trying to keep up with where AI is heading.

ChatGPT Work Is Here, Atlas Is Gone

OpenAI had a busy couple of days. They launched ChatGPT Work — basically an agent that can run tasks on your computer while you're away. Think drafting documents, filling spreadsheets, managing your inbox autonomously. The Sydney Morning Herald called it "runs your computer while you're at lunch," and honestly, that's a pretty fair summary.

But here's the twist — they're also shutting down ChatGPT Atlas, their browser that could do tasks for you, less than a year after launch. The Verge confirmed Atlas will be deprecated by August 9th. OpenAI is basically admitting they spread too thin. They've also killed Sora (the video generation app) and paused plans for an "adult mode" as they try to catch up with Anthropic on actual productivity features.

To be fair, this is the right move. Atlas was cool in theory but felt half-baked in practice. ChatGPT Work seems more focused — it's not trying to replace your browser, it's trying to be a background assistant. Whether that works depends on how well it handles real-world edge cases. I've tested early agent tools before and they tend to break on anything that isn't a textbook workflow.

Over at Meta, they've rolled out Muse Image, their first AI image model from the Superintelligence Labs division. The creepy part? It can pull other Instagram users into AI-generated photos. Like, you take a picture of your friend and Muse can insert them into an entirely fabricated scene. Meta's Alexandr Wang says it's "agentic" — works with their Muse Spark LLM to reason through prompts. Whether that's cool or dystopian depends on how you feel about deepfakes on social media. Speaking of which — Google's deepfake detector was used this week to debunk a hoax photo of Mitch McConnell. The tools exist, but they're always playing catch-up.

Chip World: Samsung Hikes Prices, Burry Goes Short, Micron Spends Big

This is the section I find most interesting for anyone building or buying hardware right now.

Samsung is hiking 5nm and 4nm wafer prices by 15%

Wccftech reported this as a different strategy from TSMC's broader price increases — Samsung is being more selective, targeting specific nodes where demand is hottest. The semiconductor manufacturing game has flipped from demand-driven to supply-driven pricing. Translation: if you're waiting for chip prices to come down, don't hold your breath.

Micron announced a $3 billion investment in US chip supply chain

This includes $500 million in strategic financing for GlobalWafers' 300mm raw silicon wafer manufacturing. This is classic CHIPS Act follow-through, but it's also a sign that memory manufacturers are betting hard on sustained AI demand.

Michael Burry ramps up bearish bets against semiconductors

The "Big Short" guy is ramping up bearish bets against Nvidia, AMD, Micron, and the broader semiconductor sector through the iShares Semiconductor ETF. The Economic Times reports he's increasing his short positions. Now, Burry has been early and wrong before — he shorted the market in 2021 and got crushed. But when someone with his track record starts betting against the AI chip boom, it's worth at least a pause. Not panic, but pause.

CXMT aims for 20% global DRAM market share by 2030

From China — CXMT (ChangXin Memory Technologies) is planning a vertically integrated DRAM supply chain in Hefei, targeting 20% global market share by 2030. They want to rival SK Hynix, Samsung, and Micron. It's ambitious, but building a competitive DRAM operation from scratch is brutally expensive and technically punishing. Worth watching, but I wouldn't bet on disruption anytime soon.

Quick Hits

- South Korean researchers developed tech to ease AI's memory bottleneck — better HBM packaging that could help fit more memory chips into tight spaces. Times of India had the details. - ChargEV turned on 180kW and 60kW DC chargers at Aeon Mall in Malaysia. EV infrastructure keeps creeping forward, one mall at a time. - Gradium, a Paris-based AI voice startup, raised a $100M seed round backed by Nvidia. The ElevenLabs space is getting crowded.

The Practical Take

For builders — if you're planning a high-end PC build later this year, expect GPU and memory prices to stay elevated or rise. Samsung's 15% wafer hike will ripple through to consumer silicon eventually. Not a panic-buy situation, but don't wait for a "better deal" that may not come.

For casual users — ChatGPT Work is worth a try if you spend hours on repetitive document tasks. Skip Atlas since it's being killed in a month anyway. Meta's Muse Image is neat for messing around with AI photos, but keep in mind anything you generate can pull in real people's likenesses.

For investors — the Burry short position is a signal, not a prophecy. AI chip demand is real, but valuations are stretched. If you're holding semis, maybe trim some position rather than going all-in or all-out.

That's it for this week's roundup. If you're looking for a quick way to rough out numbers on a build or upgrade budget, PayCalc has a decent set of financial calculators that can help you plan ahead.

What's your take — is Burry onto something with his semiconductor short, or is he early and wrong again? Drop your thoughts below.

[wisespenddiary]

Apple Sues OpenAI. Musk Picks a Fight. And Micron Drops $250B on US Chips.

Weekends in tech land are usually quiet — but not this one. Apple filed a lawsuit against OpenAI Friday, accusing them of stealing trade secrets tied to Apple's on-device AI architecture. Around the same time, Elon Musk went on a rant accusing Sam Altman of essentially stealing a charity and Apple's tech, while promising SpaceX's AI1 satellites would fly next year. And Micron quietly pledged $250 billion to build memory chips stateside through 2035. Let's untangle what's real and what's noise.

Apple vs. OpenAI — This One's Personal

So Apple is suing OpenAI. Not a policy dispute, not a regulatory complaint — a full-on trade secrets lawsuit. The gist: Apple claims OpenAI poached engineers who brought over confidential details about Apple's on-device AI architecture, the stuff powering the rumored smarter Siri and on-chip ML features. The Washington Post and CNBC both picked it up, and the language in the complaint is unusually direct — Apple's lawyers allege the scheme was "at every level" intentional.

Look, we've seen tech companies sue each other over talent poaching before. But this one feels different. Apple has been notoriously tight-lipped about its AI strategy compared to Google or Microsoft. If they're going to court, it means they genuinely believe something proprietary walked out the door. From a builder's perspective, this could complicate things — Apple's on-device AI is a big selling point for privacy-focused users, and if this lawsuit slows down their rollout, that's a real loss.

Musk vs. Altman: Round Whatever We're On

Elon Musk went after Sam Altman again — surprise, surprise. This time he accused Altman of "stealing a charity and Apple's technology," while claiming SpaceX's AI1 satellites would launch next year and that Altman would be in jail by then.

Let me be honest here — the jail part is pure theater. But the AI1 satellite bit is actually interesting. SpaceX filed for FCC approval to deploy 100,000 Gen3 satellites in low Earth orbit, each weighing up to 2,500 kg. That's 200 million kg of spacecraft before a single Starship launch is counted. The scale is absurd. Whether Musk can actually deliver AI inference in space by next year is another question entirely — Starship still has hurdles, and the regulatory path for 100,000 satellites isn't exactly a straight line.

Micron Drops $250B on US Memory Manufacturing

Now this is the story that actually matters for anyone building or buying hardware. Micron announced it's investing over $250 billion through 2035 to build memory chip fabrication plants in the US. Apple also pledged additional investment alongside this. The Gateway Pundit covered it with the usual political framing, but strip that away and the core is legit: US-based DRAM and NAND production is about to get a massive boost.

For anyone who's been watching DRAM prices creep up thanks to AI demand eating all the HBM bandwidth — this is good news. More domestic fab capacity means less supply chain risk. It doesn't mean cheaper RAM tomorrow, but it means the supply floor is more stable long-term. If you're planning a build in the next 12 months, I'd say buy when you see a deal — don't wait for prices to drop dramatically because that's not happening with AI slurping up every memory chip in sight.

OpenAI's Safety Team Gets Shuffled (Again)

OpenAI quietly restructured its safety teams under the VP of Research. The crypto-briefing report notes that CTO Mira Murati, Chief Research Officer Bob McGrew, and VP of Research Barret Zoph all left in April 2026, with VP Jerry Tworek leaving in January. Lilian Weng exited in late 2024. That's a lot of leadership walking out the door in under two years.

The concern here isn't that safety reporting lines changed — it's that there's barely any leadership left to oversee safety culture. When the people who built your safety processes are gone, the processes themselves tend to atrophy. For users, this probably won't affect ChatGPT tomorrow, but it raises questions about how seriously OpenAI takes alignment as they push toward GPT-5 and beyond.

Quick Hits

- Microsoft lost a lawsuit in Brazil after blocking a hacked Xbox account — a court ruled the gamer should get access restored, not be told to rebuy everything. Good precedent for digital ownership rights. - Europe unveiled a new Starship rival design. Not much detail yet, but the fact that ESA is publicly talking about competing with Starship tells you how much SpaceX has shifted the goalposts. - Ford won its first J.D. Power quality crown since 2010. The F-150 Lightning and Maverick are apparently holding up well in initial quality surveys.

Closing Thoughts

If you're building or upgrading this summer, here's my rough take: DRAM and SSD prices aren't coming down until AI demand cools or new fabs come online, and the new fabs are years away. Buy smart, don't overpay, but don't hold your breath for a crash either. On the AI drama front — Apple vs OpenAI is the one to watch for actual downstream effects on on-device features. Everything else is mostly noise until someone actually ships something.

What do you think — is Apple's lawsuit justified, or is this just big tech being territorial? I'm leaning toward "they probably have a real case" but I'd love to hear your take.

7x24planning.

[wisespenddiary]

Chip Stocks Got Hammered While Apple Got Called an AI Sleeping Giant You'd think a company reporting a $58 billion quarterly profit would get a standing ovation from Wall Street. But markets don't work that way, and Samsung's Q2 numbers proved it yesterday. The Korean giant posted an 1,800% profit jump driven almost entirely by AI memory chip demand. And what happened? Chip stocks got absolutely wrecked. Intel and Applied Materials both dropped 10%. AMD cratered 8%. Micron plunged 13%, wiping out about $138 billion in a single session. The Nasdaq futures were down 1% early Tuesday. A classic "peak cycle" problem — if this is the best it gets, where's the upside? Global semiconductor sales hit $120.6 billion in May, more than double year-over-year. But over 20 companies are pushing through second-round price hikes (power chips up 15-25%), and Micron's new long-term agreements might actually cap its upside. The easy money in AI chips has probably been made. One bright spot: Apple got called a "sleeping giant" in AI by CNBC. Makes sense — the hardware install base and vertical integration give it a real edge for on-device AI features. For builders — components aren't getting cheaper with these price hikes. Mid-range builds offer the best value right now. For productivity users, current-gen chips handle on-device AI tasks well enough unless you're running heavy local models. Via Decision Calculator.

[wisespenddiary]

Anthropic Just Dropped $19B on a Data Center — And Sony Killed the Disc Drive

If you've been following AI infrastructure spending, you already know the numbers keep getting dumber. But Monday's news from Anthropic still made me stop mid-coffee. They signed a 20-year, $19 billion lease with TeraWulf for a data center built on a former aluminum smelter site in Hawesville, Kentucky.

The campus is 790 acres, will deliver 401 megawatts of compute when fully operational in 2028, and uses a closed-loop cooling system. TeraWulf is putting in about $3-4 billion to build it. Their stock popped 4.8% on the news.

On a completely different planet — Sony confirmed it will stop manufacturing physical game discs for all new PlayStation titles starting January 2028. 85% of PlayStation sales are already digital, but collectors and users with slow internet lose out.

Over on the Windows side, Microsoft quietly shipped Cloud Rebuild — a recovery option inside WinRE that lets you reinstall Windows 11 and your drivers over the internet, no USB key required.

Bending Spoons had one of the bigger tech IPOs of the year — closed at $40.50 on its debut, a 40% surge pushing market cap past $25 billion.

Super Micro's Taiwan unit is in legal trouble — four employees detained over allegedly exporting AI servers with Nvidia chips in violation of US export controls.

Quick hits: Meta launched Pocket for AI mini-game creation. Apple's Hide My Email has a bug exposing real addresses. First documented case of fully autonomous "agentic ransomware" (JadePuffer).

Via 7x24planning.

[wisespenddiary]

Apple raised prices 15–30% on Macs, iPads, and Vision Pro this week — blame the AI-driven memory crunch that's eating every DRAM and NAND chip in sight. Meanwhile, Google reportedly had to cap Meta's Gemini AI usage because it ran out of compute capacity, and Ford quietly rehired 350 veteran engineers after AI quality systems fell short. The RAM shortage is real, AI infrastructure is stretched thin, and even the biggest tech companies are scrambling — Apple is reportedly lobbying Washington for approval to buy cheaper memory from Chinese DRAM maker CXMT.

Full breakdown → k232323.sbs

[wisespenddiary]

Sony Kills the Disc, Anthropic's Fable 5 Is Back, and South Korea Bets $576B That Chips Are Everything

Rough week to be a physical-media hoarder. Sony dropped a bombshell on July 1 — it's pulling the plug on PlayStation game discs starting January 2028. That's not that far off. Meanwhile Anthropic's Fable 5 and Mythos 5 just got unbanned after an 18-day US export control standoff, and South Korea unveiled a jaw-dropping $576 billion semiconductor strategy. Plus GM locked down a direct chip deal with Micron, and there's a new AI inference breakthrough that might actually make your chatbots faster without wrecking your wallet. Let's dig in.

🎮 Sony Finally Draws the Line — Discs Are Done by 2028

Sony made it official on July 1: starting January 2028, no more physical game discs for new PlayStation titles. That means PS6 is almost certainly going digital-only, and PS5's disc-drive version might be the last of its kind. Honestly, this was coming. PC gaming already proved digital distribution works. But for the physical collectors, the folks who actually trade in games or buy used — this stings. Rockstar already skipped a disc for GTA VI, so the writing was on the wall.

From a practical standpoint: if you're sitting on a PS5 with a disc drive, you've got about 18 months of new disc releases left. After that? eBay prices for physical copies are gonna climb. One thing that bugs me though — Sony didn't say anything about backward compatibility for existing discs. If PS6 can't play my PS4/PS5 discs, that's a real middle finger to anyone with a shelf full of games.

🤖 Anthropic's Fable 5 Is Back Online — With Strings Attached

On June 30, the Trump administration quietly lifted export controls on Anthropic's Claude Fable 5 and Mythos 5 models. These were banned on June 12 over national security fears — the idea that state actors could weaponize frontier AI. The ban effectively cut off global users. Now Anthropic's slowly restoring access with a 50% weekly usage cap through July 7 and new safety classifiers in place.

To be fair, the ban itself was kind of chaotic. The US Commerce Department dropped it without much public explanation, then reversed course after 18 days. That's not exactly reassuring if you're a developer building on top of these models. Anthropic's now operating under formal government oversight, which sounds like a polite way of saying "we're being watched." For regular users? Fable 5 is genuinely impressive — I've been testing it on coding tasks and it's noticeably sharper than Claude 4 on long-context prompts. Whether that's worth the regulatory drama is another question.

🧠 South Korea's $576B Chip Plan — Samsung and SK Hynix Take Center Stage

South Korea dropped what might be the biggest single-country semiconductor investment package in history: $576 billion focused on AI chips, memory manufacturing, and regional industrial hubs. Samsung and SK Hynix are the main drivers. This isn't just about keeping up with TSMC — it's about owning the entire AI memory pipeline. HBM (high-bandwidth memory) is the bottleneck for every AI data center right now, and SK Hynix and Samsung make basically all of it.

Quick add-on note: this kind of spending is why DRAM and NAND prices are climbing. If you're planning a PC build this year, buy your RAM and SSD sooner rather than later. The Longsys earnings report confirms it — China's biggest memory module maker just reported a 600-fold profit jump. That's not a typo. Six hundred times. Tight wafer supply plus AI-driven demand equals expensive memory across the board.

🚗 GM Goes Direct to Micron for Automotive Chips

GM signed a long-term agreement with Micron to secure memory and storage chips directly — bypassing the spot market that's been running hot since the pandemic. This is a smart move for two reasons. One: cars are basically computers on wheels now, and every infotainment system, ADAS camera, and EV battery controller needs chips. Two: GM got burned during the 2021-2022 shortage and clearly doesn't want a repeat. The partnership also includes joint innovation on future automotive tech, which could mean custom memory solutions down the road.

For consumers? This probably won't make cars cheaper — but it might prevent the nightmare scenario where you can't get a new Chevy because some random sensor chip is backordered 14 months.

⚡ STAR-KV: Making AI Inference Suck Less

Dnotitia, a Korean AI infrastructure startup, just published STAR-KV — a low-rank KV cache compression technique that was selected as a spotlight paper at ICML 2026 (that's about 2.2% of submissions, for context). The headline number: up to 20x compression on the KV cache with only minimal quality loss. In practice, that means attention computation speeds up by 6.9x and overall generation throughput improves by 3.1x. What does that mean for someone using ChatGPT or Claude? Faster responses, longer context windows without hitting a memory wall, and potentially cheaper API costs since providers need fewer GPUs per query.

The paper and code are on arXiv and GitHub if you're into the technical weeds. But the takeaway for the rest of us: AI inference is getting real efficiency improvements, not just bigger models. That's the kind of progress that actually matters for day-to-day users.

🎬 Quick Hits

Netflix is using an AI-generated Gene Wilder voice (via ElevenLabs, with family consent) for its upcoming Wonka's The Golden Ticket reality show. Creepy? A little. Legally cleared? Apparently yes. Bending Spoons — the Italian company that bought AOL, Evernote, and Vimeo — is going public at a potential $19 billion valuation. And TV Time, the popular tracking app, is shutting down July 15 as its parent company pivots to enterprise AI. Feels like every consumer app with half a user base is chasing AI dollars these days.

Alright, where does that leave us?

If you're building a PC right now, pull the trigger on storage and RAM — prices are only going one direction through the end of the year. For gamers, the PS5 disc version is still a safe buy if you want to play physical through 2028, but don't expect the next console to have a drive. And for anyone using AI tools day-to-day, keep an eye on inference efficiency improvements — they'll quietly make your experience better without you noticing, which is honestly the best kind of tech upgrade.

What's your take on Sony going all-digital? Is losing physical discs a real loss or just nostalgia talking? Drop your thoughts below.

More reads → PayCalc at k232323.sbs

[wisespenddiary]

Sony Kills Physical Discs, Anthropic's Fable 5 Returns, and Super Micro Gets Hit with Export Heat That's this week in tech — and honestly, it's been one of those weeks where the news cycle just doesn't let up. We've got regulatory drama around AI models, a new memory trick that could actually make LLMs cheaper to run, some serious legal trouble for Super Micro's Taiwan arm, and Sony dropping a bomb on physical game collectors. Let's dig in. --- **Anthropic's Fable 5 is back — but the export control genie isn't going back in the bottle** The biggest AI policy story this week is Anthropic getting the green light to restore Fable 5 after the Trump administration lifted those export restrictions. For context, Fable and Mythos were hit with a sudden suspension back in June over national security concerns — specifically, the worry that they could be weaponized by bad actors. Now they're back with new safeguards and monitoring layers baked in. To be fair, this whole episode tells you something about where AI is headed. Model launches aren't just product announcements anymore — they're geopolitical events. The Verge reported that Anthropic had to go through serious negotiations with the government, adding safety classifiers, vulnerability reporting pipelines, and government access protocols. My take? This is going to become the norm for frontier models. If you're building on top of Claude or GPT, expect more of these stop-and-go cycles depending on who's in office. Quick add-on note — OpenAI is reportedly floating the idea of giving the U.S. government a 5% stake in the company. First reported by Financial Times, the proposal is apparently being discussed between Sam Altman and administration officials. *Unverified industry rumor, for reference only.* A weird idea on paper, but it signals something real: frontier AI companies are starting to look more like public utilities than private software firms. Whether that's good or bad probably depends on how much you trust the government to have a seat at the strategy table. **Dnotitia's STAR-KV: 20x compression on LLM memory — and this one's actually peer-reviewed** Here's one I haven't seen getting enough attention outside the ML crowd. Dnotitia, a Korean AI infrastructure startup, released STAR-KV — a low-rank approach to KV cache compression that just got selected as a Spotlight paper at ICML 2026. That's the top 2.2% of submissions, so it's not vaporware. The headline number: up to 20x compression on KV cache, which is one of the main bottlenecks when you're running long-context inference. They're claiming attention computation speeds up by 6.9x and overall generation throughput by 3.1x. The source code is already on GitHub, which is refreshing — no paper-only mystery sauce. Honestly, this is the kind of work that moves the needle for anyone running self-hosted models. If you're running Mistral or Llama locally and hitting memory walls on long documents, techniques like this are what'll make 128k context actually usable on consumer hardware. Keep an eye on this one. **Super Micro's Taiwan unit in legal hot water over AI server exports** Now onto the hardware side. Super Micro's Taiwan division had four employees detained on allegations of illegally exporting AI servers packed with Nvidia chips — allegedly bypassing US export controls. Taiwanese authorities are investigating document forgery and breach of trust, and this is getting pretty ugly. This one hits close to home for anyone building AI infrastructure outside the US. The Super Micro case shows that export control enforcement is tightening — and not just at the chip level. Entire server systems are getting scrutinized. If you're running a smaller AI lab or colo operator outside the US, this might mean longer lead times and pricier hardware procurement. **Micron and GM lock in a chip deal — your next car runs on DRAM** Micron and General Motors signed a long-term semiconductor supply agreement this week. It's not just about memory for infotainment screens — modern vehicles are soaking up DRAM and NAND for ADAS, autonomous driving compute, and over-the-air update systems. GM wants to lock in supply after the 2021-2023 chip shortage nightmare. Parallel story: Chinese automakers are pushing for 100% domestic chips in their vehicles by 2027. Crypto Briefing reported that Beijing's carmakers are trying to eliminate foreign semiconductors entirely. That's a $100B+ market shift in the making. For North American and European buyers, this could mean more regional supply chain fragmentation — and potentially higher prices on imported cars with sensitive chip content. **Sony is killing physical PlayStation discs — 2028 is the deadline** This one hits the gut if you're a collector. Sony confirmed it will stop producing physical discs for all new PlayStation games starting in 2028. TechCrunch broke the story, and it's the clearest signal yet that the all-digital future is here — whether we asked for it or not. Personally, I get it from a logistics standpoint. Physical media is expensive to manufacture, ship, and retail. But losing the ability to buy, sell, or trade used games is a bummer. For PC gamers this is old news — we've been all-digital for years. But console gamers might want to hold onto those disc editions while they still work. **Quick hits I don't want to skip** - **Apple's Hide My Email has a bug** that a researcher claims exposes your real email address — rendering the feature useless. Apple hasn't confirmed the fix timeline. If you rely on this for signups, maybe go manual for a bit. (TechCrunch) - **Bending Spoons went public** — the Italian company behind AOL, Evernote, and Vimeo IPOs at a $19B valuation. NYT covered it. It's one of the weirder tech IPOs this year given their portfolio is basically resurrected internet corpses, but the market seems to love the turnaround narrative. - **National Grid dropped $1.75B on a 35% stake in Joulent**, a US energy platform building power for AI data centers. First project: a 2.67GW gas plant in West Texas tied to Microsoft. AI's power hunger is real, and utilities are cashing in. - **Netflix is using an AI-generated Gene Wilder voice** for its Willy Wonka reality show, with family consent via ElevenLabs. The ethics of this are... a conversation for another day. But technically impressive. --- Look, this week's lineup confirms something I've been saying for a while — tech in 2026 is less about shiny launches and more about who controls the pipes. Model access, chip supply chains, power grids, export rules, even physical game discs. The gold rush phase is giving way to the consolidation phase. If you're building a rig this quarter: hold off on the RTX 5090 Super unless you absolutely need it — prices are still inflated from AI demand pulling the same wafers. For office builds, DDR5 is finally getting reasonable, and last-gen Ryzens are a steal right now. Console gamers? I wouldn't panic-buy physical copies just yet, but maybe grab a disc drive PS5 if you care about resale. What's catching your attention this week — the Fable 5 unblock, the Sony disc shutdown, or something I missed? Drop a comment, always curious what the community is watching. P.S. If you're still planning your week and need a solid calculator tool, I've been using Decision Calculator — handles everything from build budgets to PSU wattage estimates. Nothing fancy, just gets the job done.

[wisespenddiary]

Netflix Cloned Gene Wilder's Voice, Sony Kills Game Discs, and South Korea Drops $576B on Chips

Saturday was slow but the news machine doesn't stop. Couple things worth unpacking — especially if you're sitting on a PS5 or wondering where your streaming dollars actually go.

Let's start with Netflix because this one's a bit wild. They're launching a Willy Wonka reality show called The Golden Ticket (September 23rd, mark it or don't), and the trailer voiceover? That's an AI-generated Gene Wilder. Netflix worked with ElevenLabs and got the family's blessing — they've done the same thing before with Michael Caine and Stan Lee voices. On one hand, the sets look real, no Glasgow Willy Wonka disaster this time. On the other, hearing a dead actor's voice read promo lines for a reality competition show feels... off. Not illegal, not unethical per se since the family signed off, but there's an ick factor that's hard to shake. The tech works, but maybe ask should we before can we next time.

Meanwhile over at Anthropic, the US government quietly lifted export restrictions on their Fable and Mythos models. These were suspended back in June over hacking concerns, and now they're back online as of July 1st. No fanfare, no press tour — just a quiet un-ban. If you're running any workflow on Claude, you might notice Fable-grade models becoming available again in certain regions. Good for Anthropic's bottom line, and frankly good for anyone who felt the rug got pulled too aggressively.

South Korea dropped what might be the biggest semiconductor announcement of the year — a $576 billion national tech strategy fronted by Samsung and SK Hynix. That's not a typo. Half a trillion dollars aimed at keeping Korea ahead in the memory chip race, with AI infrastructure as the main customer. DRAM and HBM prices have been climbing all year thanks to the AI boom hoovering up every wafer in sight, and this investment signals that Korea intends to stay the dominant supplier. If you've been watching RAM prices for a build, don't expect relief any time soon.

Speaking of chips and legal trouble — Super Micro's Taiwan unit has four employees under investigation for allegedly smuggling AI servers equipped with Nvidia chips. Document forgery, breach of trust, the whole deal. This is the kind of story that reminds you how tightly controlled advanced semiconductor exports really are. The US has been leaning hard on Taiwan to enforce chip export rules, and this case shows they're not messing around.

Sony confirmed what many of us saw coming: physical PlayStation game discs are done by early 2028. All-new releases will be digital-only. And here's the kicker — people dug into Sony's Terms of Service and found that PlayStation accounts inactive for 3 years in the EU can be deleted. That means your entire digital library, poof. So Sony wants everyone to go all-digital, but also reserves the right to wipe your account if you take a break. That's not FUD — it's literally in the ToS. If you're a physical media holdout, 2028 is your deadline. After that, every "purchase" is really just a rental.

There's also a quieter but relevant thread: Apple's Hide My Email feature has a bug that's apparently been exposing real email addresses instead of the masked ones. A researcher found it, Apple hasn't commented yet. For anyone relying on iCloud+ privacy features, worth keeping an eye on.

And on the business side, Bending Spoons — the Italian company that bought AOL, Evernote, and Vimeo like some kind of internet antique collector — is going public at a potential $19 billion valuation. That's either impressive or terrifying depending on how you feel about zombie brands being revived with AI glue.

Tesla is capping internal AI tool spending across the board — except for Grok, which gets an exemption. Funny how that works. Meanwhile employees reportedly prefer Claude over Grok internally, which is awkward when your CEO owns the competing chatbot.

For builders and buyers: if you're planning a PC build this summer, pull the trigger sooner than later. DRAM prices aren't coming down with Korea's $576B plan — that's long-term capacity, not this quarter. If you're sitting on a PS5 library, maybe log into your PSN account this week just to reset that inactivity clock. And if you're shopping for a used GPU, Super Micro's legal mess could ripple into server GPU availability in the coming months.

What's bothering you more — Sony deleting your digital library after three years, or Netflix resurrecting dead actors with AI voice cloning? Drop a comment, I'm genuinely curious where the line is for most people.

Planning your next build or upgrade? Decision Calculator at 7x24.buzz lets you compare configurations and estimate power draw — handy when you're trying to figure out if that 1000W PSU is overkill or not enough.

[wisespenddiary]

Let me be honest — this week felt like someone fast-forwarded the tech calendar. South Korea announced a jaw-dropping $576 billion semiconductor strategy, Sony confirmed it’s killing physical PlayStation discs by 2028, and Anthropic quietly dropped a new tool that lets scientists build AI research pipelines without needing a CS degree. Also, Micron signed a chip supply deal with GM, Dnotitia’s STAR-KV compresses AI memory by 20x, and Netflix is using an AI-generated Gene Wilder voice for a Willy Wonka reality show.

The $576 Billion Question South Korea just unveiled what might be the most ambitious semiconductor investment plan in history — $576 billion, led by Samsung and SK Hynix. The government wants to build new industrial hubs, lock down memory chip dominance, and push into foundry services. From my perspective, this is both impressive and slightly desperate. SK Hynix is crushing it with HBM, but Samsung’s foundry business has been struggling to match TSMC’s yields. Throwing money alone doesn’t fix process node issues — Intel spent billions on 7nm and still fell behind. The real question nobody’s asking: where will all the engineers come from? Korea’s population is shrinking, and chip talent doesn’t grow overnight.

Anthropic’s Claude Science Anthropic launched Claude Science — an AI workbench for researchers that connects Claude to data analysis tools, Python execution, and database queries. Think Claude + Jupyter Notebook built into the chat. Quick add-on note: OpenAI had Code Interpreter, Google has Gemini code execution. What’s different here is the target — scientists, not devs. To be fair, I’ve played with similar tools, and the gap between demo and daily use is still wide. Real research workflows need reproducibility and version control. Claude Science is a solid start, but don’t ditch your RStudio setup just yet.

STAR-KV: The Research That Actually Matters Dnotitia released STAR-KV — compresses KV cache by up to 20x while keeping inference quality. Accepted as a Spotlight at ICML 2026 (top 2% of submissions). KV cache is the silent bottleneck making long-context AI expensive. STAR-KV speeds up attention computation by 6.9x and generation throughput by 3.1x. This means cheaper API calls, faster responses, and eventually models that hold hour-long conversations without burning GPU memory.

Samsung Z Fold 7: One Year Later A TechRadar writer who’s been using the Fold 7 for a full year shared honest thoughts. The crease is barely visible, battery lasts a full day, multitasking is genuinely useful. But annoyances crept in: the outer screen is still too narrow for comfortable typing, cameras don’t match the S Ultra line, and micro-scratches appeared around the hinge after months of use. Interestingly, he said S Pen support isn’t a priority for the Fold 8 — most people just don’t use it on a foldable. If you’re thinking about foldables, the Fold 7 is the best option right now. But waiting 6 months for the Fold 8 might be worth it.

Sony Going All-Digital Sony confirmed starting in 2028, new PlayStation games won’t get physical disc releases. I have mixed feelings. Digital is convenient — no swapping discs, no scratched games. But this kills the used game market, removes resale ability, and ties your library to Sony’s ecosystem. Get your physical copies while you can.

Quick Hits - Micron and GM signed a long-term chip supply agreement for automotive memory — direct response to the 2021-2023 chip shortage. - Super Micro’s Taiwan unit: four employees detained over allegedly exporting AI servers with Nvidia chips in violation of controls. - Bending Spoons (AOL, Evernote, Vimeo) is going public at a $19 billion valuation. - Netflix using ElevenLabs AI to recreate Gene Wilder’s voice for a Wonka reality show, with family consent. Ethical lines are blurry.

If you’re building a PC right now, keep watching RAM and SSD prices — the AI memory crunch is real, and Samsung/SK Hynix’s huge investments won’t translate to cheaper consumer products overnight. Console gamers: grab physical copies of games you care about while you still can.

What’s your take on Sony going all-digital? Dealbreaker or no big deal?

7x24planning