<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Tomorrow Today — Technology, capital and applied science</title><description>Independent technology journalism covering artificial intelligence, capital markets, education, semiconductors, robotics and healthcare.</description><link>https://tomorrow-today.pages.dev/</link><language>en-US</language><item><title>Anthropic Sets a $70–$78 IPO Range Ahead of Its Nasdaq Debut</title><link>https://tomorrow-today.pages.dev/money/anthropic-ipo-price-range/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/money/anthropic-ipo-price-range/</guid><description>The Claude maker plans to sell 825 million Class A shares under the ticker ANT, with pricing expected Nov. 18 and trading due to begin the following day.</description><pubDate>Tue, 17 Nov 2026 14:00:00 GMT</pubDate><category>money</category><category>Anthropic</category><category>IPO</category><category>Nasdaq</category><category>Capital Markets</category><category>AI Compute</category><author>Kshithij Anand Belman</author></item><item><title>Gemini 4 Argon Was Built for the Long Task. Enterprises Are Still Testing the Trust.</title><link>https://tomorrow-today.pages.dev/ai/gemini-4-argon-long-horizon/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/ai/gemini-4-argon-long-horizon/</guid><description>Google&apos;s frontier model has been generally available since October, pairing a one-million-token context window with aggressive pricing and a staged safety rollout.</description><pubDate>Tue, 17 Nov 2026 13:00:00 GMT</pubDate><category>ai</category><category>Google</category><category>Gemini 4 Argon</category><category>DeepMind</category><category>Enterprise AI</category><category>Long-Horizon Tasks</category><author>Santhosh Kumar</author></item><item><title>Claude Fable 5.2 Arrives With Faster Output, Lower Prices and a Split Safety Strategy</title><link>https://tomorrow-today.pages.dev/ai/claude-fable-5-2-release/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/ai/claude-fable-5-2-release/</guid><description>Anthropic&apos;s new frontier release ships as two models: a broadly available Fable with hard cyber and biology guardrails, and a vetted-access Mythos for sensitive work.</description><pubDate>Tue, 17 Nov 2026 12:00:00 GMT</pubDate><category>ai</category><category>Anthropic</category><category>Claude</category><category>Fable 5.2</category><category>Mythos 5.2</category><category>Foundation Models</category><category>Pricing</category><author>Santhosh Kumar</author></item><item><title>Honor Is the Latest Phone Maker to Bet Its Future Includes Robots</title><link>https://tomorrow-today.pages.dev/robotics/honor-robotics-subsidiary/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/robotics/honor-robotics-subsidiary/</guid><description>The smartphone company has launched a robotics subsidiary, following a path Chinese consumer-electronics firms are increasingly treating as logical.</description><pubDate>Tue, 17 Nov 2026 10:30:00 GMT</pubDate><category>robotics</category><category>Honor</category><category>Humanoid Robots</category><category>China</category><category>Consumer Electronics</category><category>Manufacturing</category><author>Kshithij Anand Belman</author></item><item><title>OpenAI Signs 457MW Compute Deal With TensorWave to Broaden Its Chip Base</title><link>https://tomorrow-today.pages.dev/ai/tensorwave-openai-457mw-amd/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/ai/tensorwave-openai-457mw-amd/</guid><description>The six-year agreement puts OpenAI capacity on AMD Instinct accelerators, a small but deliberate step away from a single-vendor supply chain.</description><pubDate>Tue, 17 Nov 2026 09:15:00 GMT</pubDate><category>ai</category><category>OpenAI</category><category>AMD</category><category>Data Centers</category><category>Compute</category><category>Inference</category><author>Kshithij Anand Belman</author></item><item><title>Google Adds Textbook-Verified Answers and Stricter Controls to Gemini for Schools</title><link>https://tomorrow-today.pages.dev/schooling/google-school-gemini-controls/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/schooling/google-school-gemini-controls/</guid><description>The updates give teachers more visibility into student AI use and tie Gemini&apos;s responses to OpenStax content in an effort to reduce hallucinated explanations.</description><pubDate>Tue, 17 Nov 2026 08:00:00 GMT</pubDate><category>schooling</category><category>Google</category><category>Gemini</category><category>Education</category><category>Classroom</category><category>AI Safety</category><author>Trisha SS Belman</author></item><item><title>Hybrid Reasoning Models: Why Developers Want Configurable Thinking Budgets</title><link>https://tomorrow-today.pages.dev/ai/claude-hybrid-reasoning-agents/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/ai/claude-hybrid-reasoning-agents/</guid><description>Frontier architectures now allow software engineers to dial thinking tokens up or down depending on task complexity and budget limits.</description><pubDate>Mon, 16 Nov 2026 14:30:00 GMT</pubDate><category>ai</category><category>Anthropic</category><category>Claude</category><category>Reasoning</category><category>Developer APIs</category><category>Software Architecture</category><author>Santhosh Kumar</author></item><item><title>LangChain Raises $300M at a $3B Valuation as Agent Tooling Matures</title><link>https://tomorrow-today.pages.dev/money/langchain-300m-series-c/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/money/langchain-300m-series-c/</guid><description>The Series C, led by Meritech with participation from Nvidia&apos;s VC arm, reflects a market treating agent observability as infrastructure.</description><pubDate>Mon, 16 Nov 2026 14:00:00 GMT</pubDate><category>money</category><category>LangChain</category><category>Funding</category><category>Venture Capital</category><category>AI Agents</category><category>Series C</category><author>Kshithij Anand Belman</author></item><item><title>FDA Clears Bracco and Avicenna.AI&apos;s CT Tool for Incidental Brain Aneurysms</title><link>https://tomorrow-today.pages.dev/healthcare/bracco-avicenna-aneurysm-fda/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/healthcare/bracco-avicenna-aneurysm-fda/</guid><description>The software flags previously undetected vascular findings in routine scans, a use case where AI augments radiologists rather than replacing reads.</description><pubDate>Mon, 16 Nov 2026 11:00:00 GMT</pubDate><category>healthcare</category><category>FDA</category><category>Radiology</category><category>Brain Aneurysm</category><category>Medical Imaging</category><category>AI</category><author>Trisha SS Belman</author></item><item><title>Cerebras Wants Developers to Program Its Wafer-Scale Chip Like Any Other Accelerator</title><link>https://tomorrow-today.pages.dev/ai/cerebras-tungsten-wafer-scale/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/ai/cerebras-tungsten-wafer-scale/</guid><description>A new language and compiler for the 900,000-core Wafer-Scale Engine tries to remove the specialist knowledge that has kept the platform niche.</description><pubDate>Mon, 16 Nov 2026 10:00:00 GMT</pubDate><category>ai</category><category>Cerebras</category><category>Compilers</category><category>Wafer-Scale</category><category>HPC</category><category>Developer Tools</category><author>Trisha SS Belman</author></item><item><title>Rapidus Weighs a One-Year Delay to 2nm Mass Production</title><link>https://tomorrow-today.pages.dev/semiconductor/rapidus-2nm-delay/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/semiconductor/rapidus-2nm-delay/</guid><description>Japan&apos;s state-backed foundry is struggling with transistor integration and yield, pushing its commercial start date toward 2029.</description><pubDate>Mon, 16 Nov 2026 07:30:00 GMT</pubDate><category>semiconductor</category><category>Rapidus</category><category>2nm</category><category>Japan</category><category>Foundry</category><category>Manufacturing</category><author>Santhosh Kumar</author></item><item><title>Exoskeleton Maker Youlong Raises $100M to Scale Factory-Floor Robots</title><link>https://tomorrow-today.pages.dev/robotics/youlong-exoskeleton-manufacturing/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/robotics/youlong-exoskeleton-manufacturing/</guid><description>The Shenzhen company&apos;s funding reflects growing demand for wearable industrial systems that augment workers rather than replace them.</description><pubDate>Sun, 15 Nov 2026 12:00:00 GMT</pubDate><category>robotics</category><category>Youlong</category><category>Exoskeletons</category><category>Manufacturing</category><category>Funding</category><category>China</category><author>Kshithij Anand Belman</author></item><item><title>Estonia&apos;s National AI Programme Reaches One in Five Upper Secondary Students</title><link>https://tomorrow-today.pages.dev/schooling/estonia-ai-leap-chatgpt/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/schooling/estonia-ai-leap-chatgpt/</guid><description>The AI Leap initiative, built with OpenAI, has become a test case for what country-wide classroom deployment looks like in practice.</description><pubDate>Sun, 15 Nov 2026 11:00:00 GMT</pubDate><category>schooling</category><category>Estonia</category><category>OpenAI</category><category>Education Policy</category><category>ChatGPT</category><category>Schools</category><author>Trisha SS Belman</author></item><item><title>The Local AI Revolution: Running 70B Parameter Models on Workstation Silicon</title><link>https://tomorrow-today.pages.dev/ai/local-llm-silicon-revolution/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/ai/local-llm-silicon-revolution/</guid><description>High-bandwidth unified memory architectures and 2-bit quantization kernels allow researchers to sever reliance on centralized cloud APIs.</description><pubDate>Sun, 15 Nov 2026 11:00:00 GMT</pubDate><category>ai</category><category>Local LLM</category><category>Hardware</category><category>Quantization</category><category>Apple Silicon</category><category>Privacy</category><author>Santhosh Kumar</author></item><item><title>Memory Shortages Are Becoming a Constraint on Everything Else</title><link>https://tomorrow-today.pages.dev/semiconductor/memory-shortage-2027-prices/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/semiconductor/memory-shortage-2027-prices/</guid><description>DRAM and HBM allocations are tightening into 2027, and manufacturers are warning that capacity, not budget, will decide who ships hardware.</description><pubDate>Sun, 15 Nov 2026 09:00:00 GMT</pubDate><category>semiconductor</category><category>Memory</category><category>DRAM</category><category>HBM</category><category>Supply Chain</category><category>Data Centers</category><author>Santhosh Kumar</author></item><item><title>Neurora&apos;s AI Test for Autism in Young Children Wins FDA Approval</title><link>https://tomorrow-today.pages.dev/healthcare/neurora-autism-fda-approval/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/healthcare/neurora-autism-fda-approval/</guid><description>The system analyzes EEG data to flag indicators in children aged three to five, offering an earlier and more objective screening path.</description><pubDate>Sat, 14 Nov 2026 15:30:00 GMT</pubDate><category>healthcare</category><category>FDA</category><category>Autism</category><category>Diagnostics</category><category>EEG</category><category>Pediatrics</category><author>Trisha SS Belman</author></item><item><title>The Compute Gap Is Now Measured in Orders of Magnitude</title><link>https://tomorrow-today.pages.dev/ai/us-china-ai-compute-gap/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/ai/us-china-ai-compute-gap/</guid><description>A new estimate puts US AI data center capacity at roughly 40 times China&apos;s, a number that says as much about electricity and packaging as about chip design.</description><pubDate>Sat, 14 Nov 2026 13:00:00 GMT</pubDate><category>ai</category><category>Policy</category><category>Data Centers</category><category>China</category><category>Export Controls</category><category>Analysis</category><author>Kshithij Anand Belman</author></item><item><title>Agile Robots Lands $500M, Signaling That Industrial Automation Is Back in Favor</title><link>https://tomorrow-today.pages.dev/money/agile-robots-500m-raise/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/money/agile-robots-500m-raise/</guid><description>The German spin-out from the DLR research institute reached an $8.5 billion valuation as investors look past humanoids to factory-floor revenue.</description><pubDate>Sat, 14 Nov 2026 10:30:00 GMT</pubDate><category>money</category><category>Agile Robots</category><category>Funding</category><category>Industrial Automation</category><category>Germany</category><category>Series</category><author>Kshithij Anand Belman</author></item><item><title>Global Startup Funding Crossed $3 Billion for the Ninth Straight Week</title><link>https://tomorrow-today.pages.dev/money/weekly-funding-3b-streak/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/money/weekly-funding-3b-streak/</guid><description>The streak is the longest of 2025 and reflects a market where a handful of large rounds now define the weekly total.</description><pubDate>Fri, 13 Nov 2026 12:00:00 GMT</pubDate><category>money</category><category>Funding</category><category>Startups</category><category>Markets</category><category>Venture Capital</category><author>Kshithij Anand Belman</author></item><item><title>Samsung and SK Hynix Are Racing for EUV Tools ASML Cannot Build Fast Enough</title><link>https://tomorrow-today.pages.dev/semiconductor/asml-euv-tool-race/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/semiconductor/asml-euv-tool-race/</guid><description>Lithography capacity has become the pacing item in advanced memory production, and the queue is measured in quarters.</description><pubDate>Fri, 13 Nov 2026 11:30:00 GMT</pubDate><category>semiconductor</category><category>ASML</category><category>EUV</category><category>Samsung</category><category>SK Hynix</category><category>Lithography</category><author>Santhosh Kumar</author></item><item><title>A UK Trial Suggests AI Tutoring Works Best With a Teacher in the Loop</title><link>https://tomorrow-today.pages.dev/schooling/ai-tutor-uk-trial-results/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/schooling/ai-tutor-uk-trial-results/</guid><description>Seven months of classroom use showed gains in engagement but reinforced that the software is a supplement, not a substitute, for instruction.</description><pubDate>Fri, 13 Nov 2026 09:30:00 GMT</pubDate><category>schooling</category><category>AI Tutors</category><category>United Kingdom</category><category>Research</category><category>EdTech</category><category>Classroom</category><author>Trisha SS Belman</author></item><item><title>A Korean Chip Merger Aims to Give Robots Better On-Device Vision</title><link>https://tomorrow-today.pages.dev/robotics/mobilint-chips-media-robotics/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/robotics/mobilint-chips-media-robotics/</guid><description>Mobilint&apos;s acquisition of video IP firm Chips&amp;Media is a bet that edge inference, not cloud connectivity, will define capable robots.</description><pubDate>Fri, 13 Nov 2026 08:45:00 GMT</pubDate><category>robotics</category><category>Mobilint</category><category>Chips&amp;Media</category><category>Edge AI</category><category>Robotics Chips</category><category>Korea</category><author>Kshithij Anand Belman</author></item><item><title>Companies Are Buying Agent Infrastructure Faster Than They Can Measure It</title><link>https://tomorrow-today.pages.dev/ai/agent-evaluation-bottleneck/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/ai/agent-evaluation-bottleneck/</guid><description>Agent platforms are raising at record valuations while buyers still lack reliable ways to tell whether a deployed agent is working, or quietly failing.</description><pubDate>Thu, 12 Nov 2026 16:00:00 GMT</pubDate><category>ai</category><category>AI Agents</category><category>Enterprise</category><category>Evaluation</category><category>Analysis</category><author>Trisha SS Belman</author></item><item><title>Heartvue&apos;s AI Cardiac MRI Analysis Clears a Key FDA Review Step</title><link>https://tomorrow-today.pages.dev/healthcare/heartvue-cardiac-mri-fda/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/healthcare/heartvue-cardiac-mri-fda/</guid><description>The company passed the agency&apos;s Predetermined Change Control Plan process, allowing it to update its model without a new submission each time.</description><pubDate>Thu, 12 Nov 2026 10:00:00 GMT</pubDate><category>healthcare</category><category>FDA</category><category>Cardiology</category><category>MRI</category><category>Medical Imaging</category><category>Regulation</category><author>Trisha SS Belman</author></item><item><title>Fractile Raises $720M to Chase Faster Inference Silicon</title><link>https://tomorrow-today.pages.dev/semiconductor/fractile-720m-inference-chips/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/semiconductor/fractile-720m-inference-chips/</guid><description>The UK chip startup is valued at roughly $7 billion after a Series C round, betting that agentic workloads reward a different architecture than training.</description><pubDate>Thu, 12 Nov 2026 08:00:00 GMT</pubDate><category>semiconductor</category><category>Fractile</category><category>Inference</category><category>Startups</category><category>Chips</category><category>United Kingdom</category><author>Santhosh Kumar</author></item><item><title>Mercor&apos;s $750M Round Prices the Market for AI Training Labor</title><link>https://tomorrow-today.pages.dev/money/mercor-750m-hiring-market/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/money/mercor-750m-hiring-market/</guid><description>The data-labelling and expert-sourcing startup reached a $10 billion valuation, reflecting how valuable human judgement remains in model development.</description><pubDate>Wed, 11 Nov 2026 16:45:00 GMT</pubDate><category>money</category><category>Mercor</category><category>Funding</category><category>Data Labeling</category><category>AI Training</category><category>Labor</category><author>Kshithij Anand Belman</author></item><item><title>TSMC Denies a Move Into 3D DRAM, but the Pressure Behind the Report Remains</title><link>https://tomorrow-today.pages.dev/semiconductor/tsmc-3d-dram-denial/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/semiconductor/tsmc-3d-dram-denial/</guid><description>The foundry says it is staying focused on logic. Its largest customer&apos;s memory problem is not going away.</description><pubDate>Wed, 11 Nov 2026 15:00:00 GMT</pubDate><category>semiconductor</category><category>TSMC</category><category>3D DRAM</category><category>Nvidia</category><category>Memory</category><category>Foundry</category><author>Santhosh Kumar</author></item><item><title>Humanoid Robots Are Entering Factories Faster Than the Evidence Says They Should</title><link>https://tomorrow-today.pages.dev/robotics/humanoid-robot-mass-production/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/robotics/humanoid-robot-mass-production/</guid><description>Mass-production announcements are arriving on schedule. Reliable, general-purpose performance in real workplaces is not.</description><pubDate>Wed, 11 Nov 2026 13:30:00 GMT</pubDate><category>robotics</category><category>Humanoid Robots</category><category>Manufacturing</category><category>Automation</category><category>Analysis</category><author>Kshithij Anand Belman</author></item><item><title>A Texas AI School Drew Scrutiny, and the Debate Is Bigger Than One Campus</title><link>https://tomorrow-today.pages.dev/schooling/alpha-school-ai-backlash/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/schooling/alpha-school-ai-backlash/</guid><description>Alpha School&apos;s model of software-led instruction and high fees has become a flashpoint in arguments over what AI in education is actually for.</description><pubDate>Tue, 10 Nov 2026 14:15:00 GMT</pubDate><category>schooling</category><category>Alpha School</category><category>Private Schools</category><category>EdTech</category><category>Texas</category><category>Policy</category><author>Trisha SS Belman</author></item><item><title>DeepSeek-R1 Pushes Open Reasoning Models Into the Mainstream</title><link>https://tomorrow-today.pages.dev/ai/deepseek-r1-open-reasoning/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/ai/deepseek-r1-open-reasoning/</guid><description>The Chinese lab&apos;s latest model matches proprietary reasoning performance on math and code benchmarks while shipping weights that anyone can run.</description><pubDate>Tue, 10 Nov 2026 14:15:00 GMT</pubDate><category>ai</category><category>DeepSeek</category><category>Open Source</category><category>Reasoning</category><category>LLMs</category><author>Santhosh Kumar</author></item><item><title>The FDA&apos;s New AI Principles Put the Burden of Proof on Deployers</title><link>https://tomorrow-today.pages.dev/healthcare/fda-genai-healthcare-comments/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/healthcare/fda-genai-healthcare-comments/</guid><description>Public comments on the agency&apos;s draft generative AI framework reveal a health system still working out who is responsible when a model is wrong.</description><pubDate>Tue, 10 Nov 2026 09:00:00 GMT</pubDate><category>healthcare</category><category>FDA</category><category>Generative AI</category><category>Regulation</category><category>Policy</category><category>Health Systems</category><author>Trisha SS Belman</author></item><item><title>OpenAI and Microsoft Release Workplace AI Agent Benchmarks</title><link>https://tomorrow-today.pages.dev/ai/ai-agents-workplace/</link><guid isPermaLink="true">https://tomorrow-today.pages.dev/ai/ai-agents-workplace/</guid><description>Two of the world&apos;s most valuable technology companies have published competing benchmarks designed to measure how effectively autonomous AI agents execute real-world knowledge work.</description><pubDate>Mon, 09 Nov 2026 08:30:00 GMT</pubDate><category>ai</category><category>OpenAI</category><category>Microsoft</category><category>Agents</category><category>Enterprise</category><author>Trisha SS Belman</author></item></channel></rss>