Tech News by InClicks

▲ Anthropic passes OpenAI at $965B valuation A valuation is the price investors put on a company's future - and in AI, that number just set a new record. On May 28, 2026, Anthropic, the maker of Claude, raised $65 billion and was valued at $965 billion , passing OpenAI to become the world's most valuable AI startup. For anyone who uses AI every day or invests in it, this is a milestone worth understanding. How Anthropic Passed OpenAI According to Axios, CNBC, and Reuters, Anthropic closed a Series H round worth $65 billion , lifting its post-money valuation to $965 billion - among the largest private tech raises in history. That figure edged past OpenAI, the long-standing leader, marking the first time Anthropic has held the title of most valuable private AI company. The round leaves it just $35 billion short of the $1 trillion mark. ▲ $65B raised, Amazon adds $5B $65 Billion - Inside the Series H Round The round was led by Altimeter, Dragoneer, Greenoaks, and Se...

▲ World's #1 chipmaker TSMC warns on power AI power efficiency refers to designing chips that do the same work using less electricity, rather than simply chasing raw speed. TSMC, the world's #1 chip foundry (a company that manufactures chips for other firms), has now declared that the real limit for AI chips is no longer performance - it is power. For anyone who uses AI services or follows chip and energy stocks, this is a meaningful signal. Why Power Became the Limit for AI Chips According to Reuters, TSMC's SVP of Business Development Kevin Zhang said the surge in electricity demand from AI has become the most important constraint shaping future chip development. From smartphones to AI data centers, most customers now want stronger performance without a sharp rise in power use. For years, chipmakers improved performance by packing in more transistors (the tiny switches inside a chip), but for power-hungry AI workloads that approach alone has hit a wall. ▲ What TSM...

▲ Qualcomm-ByteDance AI chip deal at a glance An ASIC is a chip custom-built to run one specific task as fast and cheaply as possible. For anyone who uses AI, this matters: ByteDance, the parent of TikTok, is buying millions of Qualcomm's custom AI chips, and the move puts a real crack in the NVIDIA GPU monopoly that has defined the AI chip market so far. Why Did ByteDance Choose Qualcomm AI Chips? According to Bloomberg, ByteDance plans to secure millions of Qualcomm data center AI chips to power its AI agent software - AI that performs tasks on a user's behalf. The interesting detail is the chip's origin: ByteDance has already designed the chip itself , and Qualcomm is helping bring that design into mass production. For Qualcomm, long known as a smartphone chip leader, this is a major customer win in the AI data center market. ▲ Millions of chips, Qualcomm stock up 7-8% Qualcomm Stock Hits an All-Time High On the news, Qualcomm shares jumped roughly 7-8% to ...

▲ Claude 4.8 launch - new AI coding and math champion Claude 4.8, also known as Claude Opus 4.8, is Anthropic's newest flagship AI model released on May 28, 2026, setting top scores on three major benchmarks simultaneously - coding, advanced mathematics, and long-document retrieval. If you follow AI model performance, these numbers are hard to ignore. What Is Claude 4.8 and What Changed? Claude 4.8 introduces three significant upgrades over its predecessor. First, Effort Control - a Low/High/Max dial that lets users choose how deeply the model reasons through a problem, preserving usage limits for tasks that need it. Second, parallel sub-task execution in Claude Code allows hundreds of tasks to run simultaneously, letting developers refactor entire codebases in one pass. Third, the model has become measurably more honest: it now says "I don't know" and acknowledges errors, with unsupported answers dropping noticeably in user testing. Fast Mode, the lighter i...

▲ Altman and Amodei walk back AI jobs apocalypse The "AI jobs apocalypse" is the prediction that artificial intelligence will eliminate half of white-collar work within one to five years. OpenAI's Sam Altman and Anthropic's Dario Amodei were the loudest voices behind it. In May 2026, both walked it back. The more revealing question is why - and the timing points squarely to one answer: both companies are racing toward $1 trillion IPOs . Altman: "I was pretty wrong" On May 26, 2026, Altman told Commonwealth Bank of Australia CEO Matt Comyn that his prediction about AI eliminating entry-level white-collar jobs was "pretty wrong" . The reversal contradicts his June 2025 podcast warning that "a lot of jobs will go away." Amodei's flip is even sharper. In May 2025 he claimed AI could erase up to 50% of white-collar jobs. This May, he reframed automation as a productivity multiplier: "multiplies work 10x" rather than des...

▲ SK Hynix joins the $1 trillion club The trillion-dollar club is the informal label for companies whose market capitalization tops $1 trillion . On May 27, 2026 , SK Hynix became the second Korean firm to clear that bar - making Korea the only country outside the US with two trillion-dollar companies. The catalyst is HBM (High Bandwidth Memory - the ultra-fast specialty memory that AI servers cannot run without). Why the $1 Trillion Milestone Matters According to CNBC and Bloomberg, SK Hynix shares surged about 11% on May 27, pushing market cap past $1 trillion and joining Samsung as Korea's second trillion-dollar firm. Year-to-date the stock is up roughly 250% , and only 16 months ago it was worth less than $100 billion . Korea's KOSPI index has surged about 86% in the same window, its best run since 1999. ▲ +250% YTD, +11% in one day HBM4 and NVIDIA: The Real Driver The engine is NVIDIA's upcoming AI accelerator platform, Vera Rubin . TrendForce and Korean...

▲ Waymo pauses 6 cities, recalls 3,791 robotaxis The Waymo recall is a mandatory software update that Waymo, the autonomous-driving subsidiary of Alphabet, pushed to all 3,791 of its 5th- and 6th-generation robotaxis (self-driving cars). The fix went live on May 12, 2026 and is the largest single-company recall in commercial autonomous-driving history. For U.S. riders, this matters for two reasons. First, Waymo runs commercial service in more than a dozen metros today, so if you live in or visit those cities, your ride availability changed this week. Second, the rollout pace of robotaxi service in your area depends on whether Waymo can close gaps like this one. Why did Waymo issue its largest-ever recall? The trigger event occurred on April 20, 2026 in San Antonio, Texas. An empty Waymo robotaxi entered a flooded road and was swept into a creek. No passengers were on board and no one was injured, but the incident exposed a clear gap: the vehicle could not recognize standin...

▲ OpenAI model cracks 80-year geometry conjecture alone AI autonomous math solving means an AI model reasoning through a problem on its own, without human hints or worked examples. On May 21, 2026 , OpenAI announced that its general-purpose reasoning model autonomously disproved a discrete geometry conjecture that had been unsolved for 80 years. A Fields Medalist publicly called the result a milestone, signaling that AI is moving beyond imitating humans and starting to add new lines to the body of human knowledge. Why AI Solving Math Is a Shock Until now, AI models learned from answers humans had already produced. Problems that no one had ever solved were effectively invisible to AI because there was no training data to copy from. OpenAI's new result flips this. The model reasoned through an open discrete geometry conjecture (the branch of math dealing with points, lines, and shapes) and disproved it on its own. ▲ Key facts behind the 80-year AI math breakthrough A Fiel...

▲ Alibaba Zhenwu M890 targets Nvidia H20 Alibaba Zhenwu M890 is a domestic AI accelerator unveiled by Alibaba's chip subsidiary T-Head. Designed for both training and inference workloads, the chip is the clearest evidence yet that China's semiconductor independence push is moving from slideware to silicon. For US readers, the consequences land in two places that matter: cloud AI pricing and the global memory and semiconductor-equipment trade. Why China's Own AI Chip Matters Now Since 2022, US export controls have blocked Nvidia's top-tier H100-class GPUs (graphics processing units, the high-end chips that power AI training) from being sold into China. Even H20 - the dialed-down chip Nvidia built specifically for the China market - was eventually cut off. Blocked imports pushed Beijing's biggest tech companies straight into mass-producing domestic alternatives. Alibaba, Huawei, and Cambricon have all moved into volume production, and the Zhenwu M890 is the la...

▲Challenging TSMC's monopoly, 2026 Memory supply priority - the power to decide who gets chips first, and at what price - has become Samsung's most potent weapon in the AI era. On May 22, 2026, Samsung chairman Lee Jae-yong flew to Taipei just 24 hours after an 18-day worker strike ended, meeting MediaTek CEO Rick Tsai directly to propose a three-part deal that could crack TSMC's manufacturing monopoly. Why Memory Became the Semiconductor Industry's New Power AI server investment surged through late 2025 and into 2026, creating an unexpected bottleneck: DRAM. Prices for high-bandwidth memory used in AI training systems roughly doubled in the six months leading up to Q1 2026, according to TrendForce market data. For chip designers like MediaTek - which ships processors for roughly half the world's Android smartphones and designs custom AI chips for Google - securing enough memory at a reasonable price is now a survival-level concern, not just a procurement iss...

▲ IDTechEx forecast, humanoid prices crash 68% by 2030 A humanoid robot payback period is the time it takes for a robot's operating output to recover its purchase cost. UK research firm IDTechEx released a May 19 report showing that, under high-utilization industrial conditions, the humanoid payback period has already collapsed to about 6 months . The finding lands directly on the desk of every operations manager weighing a humanoid procurement, and on every worker wondering when automation hits their role. Three years of stalled decisions, and why the answer arrived now Manufacturing and logistics operators have spent the last three years stalling humanoid procurement, with one question dominating every meeting: when does the robot actually pay for itself? IDTechEx built its answer on real deployment data from Boston Dynamics Atlas, Figure 02, and Agility Digit running in commercial settings, then layered hardware cost trajectories on top. The conclusion is direct. Under ...

▲ Google AI Ultra 60% cut + Gemini Spark unveiled AI Ultra is Google's top-tier AI subscription, bundling its latest Gemini models with autonomous assistant features. If you're paying for ChatGPT every month, the announcement Google made yesterday is worth a pause. At Google I/O 2026 (May 19), held at the Shoreline Amphitheatre in Mountain View, CEO Sundar Pichai unveiled a major restructuring: AI Ultra now starts at $100 per month, down from $250 - a 60% cut . Google simultaneously lowered the previous $250 top tier to $200, keeping the same features. In short, AI Ultra is now split into two price tiers: a new $100 entry option and a $200 top tier (formerly $250). Why Google halved its AI plan The cut isn't a routine discount - it reflects a power shift in the AI market. Rival Anthropic is reportedly in talks at a valuation as high as $950B , eclipsing OpenAI's most recent reported valuation of $825B . With the duopoly being challenged, Google is using price a...