Google's Gemini 3 Deep Think is demonstrating exceptional capabilities, achieving a 3455 Elo rating on Codeforces, a score that places it in Legendary Grandmaster territory, surpassing all but seven human programmers. This upgraded model, launched on February 12, 2026, also scored an impressive 84.6% on the ARC-AGI-2 benchmark, outperforming competitors like Claude Opus 4.6 and GPT-5.2. Designed for complex science, research, and engineering challenges, it is available to Google AI Ultra subscribers and via Vertex AI.
Early testers, including Lisa Carbone from Rutgers University, have used Deep Think to identify logical flaws in math papers, while Duke University's Wang Lab optimized crystal growth for semiconductor materials. However, the broader AI boom is causing significant market volatility, with billions wiped from Big Tech and Bitcoin. Major companies like Amazon, Microsoft, Alphabet, and Meta Platforms are facing investor scrutiny due to hundreds of billions in AI data center capital spending, leading to concerns about returns.
The market is now shifting from broad AI gains to discerning winners and losers, with industries such as software, legal services, and freight brokerage facing disruption from AI automation. Meanwhile, Chinese AI company Minimax has released its M2.5 large language model, which achieved 80.2% on the SWE-Bench Verified coding benchmark, surpassing GPT-5.2 and Gemini 3 Pro. This model is also claimed to be 37% faster for complex tasks at a cost of just $1 per hour, highlighting China's rapid advancements.
In the UK, BT Business is committing to AI training for all 11,000 employees, creating new AI apprentice and graduate roles with support from Avanade and Microsoft. London-based Aeternum Tech LTD is expanding its AI-enabled enterprise system services globally, focusing on areas like business intelligence and supply chain optimization. Furthermore, Strategy, formerly MicroStrategy, introduced an AI platform that integrates context-aware AI across all business applications, ensuring consistent intelligence and data security.
The finance industry is also undergoing a rapid transformation, with AI offering faster solutions for tasks from algorithmic trading to fraud prevention, attracting more early-stage investments. On the political front, Tech for Campaigns, a Democratic-aligned group, is launching an "AI Campaign Lab" to help Democrats effectively leverage AI in elections, funding experiments to shape candidate appearances in AI chatbot results and ensure they keep pace with AI's impact on campaigning.
Key Takeaways
- Google's Gemini 3 Deep Think achieved a 3455 Elo rating on Codeforces and an 84.6% verified score on ARC-AGI-2, outperforming competitors like Claude Opus 4.6 and GPT-5.2.
- The AI boom is causing significant market volatility, with billions wiped from Big Tech and Bitcoin, and major companies like Amazon, Microsoft, Alphabet, and Meta Platforms facing scrutiny over AI capital spending.
- AI is disrupting industries such as software, legal services, freight brokerage, and finance, leading to a market shift from broad gains to picking specific winners and losers.
- Chinese AI company Minimax released its M2.5 large language model, which scored 80.2% on the SWE-Bench Verified coding benchmark, surpassing GPT-5.2 and Gemini 3 Pro, and costs $1 per hour.
- BT Business is providing AI training to all 11,000 employees and creating new AI apprentice and graduate roles, with training support from Avanade and Microsoft.
- Aeternum Tech LTD is expanding its global AI-enabled enterprise platform services, focusing on areas like AI-driven business intelligence and supply chain optimization.
- Strategy (formerly MicroStrategy) launched an AI platform that integrates context-aware AI across all business applications with a unified semantic layer and enterprise-grade data security.
- Tech for Campaigns is establishing an "AI Campaign Lab" to help Democrats utilize AI in elections, funding experiments ranging from $50,000 to $150,000.
- AI is transforming the finance industry by optimizing algorithmic trading, credit scoring, and fraud prevention, attracting increased early-stage investments.
- Investment opportunities are increasingly seen in overseas markets, particularly China, due to its advantage in AI and predicted boom in chip IPOs.
Google Gemini 3 Deep Think Outperforms Most Human Coders
Google's Gemini 3 Deep Think achieved an impressive 3455 Elo rating on Codeforces. This score places it in Legendary Grandmaster territory, surpassing all but seven human programmers active recently. Google DeepMind tested the AI in "no tools" mode, meaning it relied solely on its internal reasoning capabilities. This demonstrates AI's rapid progress in solving complex, time-constrained algorithmic problems. The enhanced mode is available to Google AI Ultra subscribers and through an early access Vertex AI program.
Google Gemini 3 Deep Think Excels in AI Benchmarks
Google DeepMind launched an upgraded Gemini 3 Deep Think, designed for modern science, research, and engineering challenges. It achieved a verified score of 84.6% on the ARC-AGI-2 benchmark, significantly outperforming competitors like Claude Opus 4.6 and GPT-5.2. The model also scored 96% on the original ARC-AGI 1 benchmark and a 3455 Elo rating on Codeforces. These results highlight its strong abilities in abstract reasoning, academic knowledge, and coding prowess. Google AI Ultra subscribers can access it in the Gemini app, and an early access program is available via Vertex AI.
Gemini 3 Deep Think Boosts Science and Engineering
Google updated Gemini 3 Deep Think on February 12, 2026, to help solve complex science, research, and engineering problems. Google AI Ultra subscribers can now use it in the Gemini app, and select professionals can access it via the Gemini API. Early testers like Lisa Carbone from Rutgers University used it to find a subtle logical flaw in a math paper. The Duke University Wang Lab also utilized Deep Think to optimize crystal growth for semiconductor materials. The model achieved high scores on benchmarks like Humanity's Last Exam (48.4%) and Codeforces (3455 Elo).
AI Boom Causes Market Swings for Bitcoin and Big Tech
Global markets are experiencing a rollercoaster ride, with billions wiped from Big Tech and Bitcoin. This volatility is deeply tied to the AI boom, causing software stocks to wobble and crypto to swing wildly. Bitcoin has halved from its peak, and massive selling has occurred in gold, US tech, and Bitcoin. An update from AI company Anthropic showed AI's potential to disrupt industries like software and legal services. Additionally, major tech companies are spending hundreds of billions on AI data centers, raising investor concerns about returns.
AI Shifts Stock Market From Gains to Losses
The stock market is seeing a shift as AI's impact moves from boosting all stocks to creating losers. While AI enthusiasm fueled the U.S. bull market, concerns about its disruptive power now rattle industries like software and legal services. Companies like Amazon, Microsoft, Alphabet, and Meta Platforms face pressure due to massive AI capital spending. Experts note a breakdown in the "monolithic AI trade," with investors now picking perceived winners and losers. The software and services index's price-to-earnings ratio recently hit its lowest in nearly three years, creating potential buying opportunities for some.
BT Business Trains All Staff in AI and Creates New Roles
BT Business, a UK broadband provider, will offer AI training to all 11,000 employees. This initiative includes new AI apprentice and graduate roles, aligning with National Apprenticeship Week. Avanade and Microsoft will provide training modules, webinars, and learning paths. John James, CEO of BT Business, stated that AI is transforming how they work and serve customers. The company is also a founding partner in the Government's "AI Skills Boost" initiative, showing its commitment to developing AI skills in the UK.
AI Threatens Freight Brokers Other Stocks Seem Safe
Artificial intelligence is now disrupting the freight brokerage industry. Companies that connect shippers with carriers, often with few physical assets, face automation from AI-driven platforms. This raises concerns about their future profits and ability to stay in business. Investors are now questioning if any asset-light business model is safe from AI's transformative power. The article suggests scrutinizing these companies and considering those with stronger business models or those that can effectively use AI rather than being replaced by it.
Aeternum Tech Expands AI Platform Services Globally
Aeternum Tech LTD, a London-based firm, announced on February 12, 2026, it is expanding its international platform development services. The company specializes in AI-enabled enterprise systems designed for optimal execution and scalability. Aeternum Tech will leverage its deep expertise in machine learning and data analytics to help businesses transform their operations. The expansion includes new development hubs and hiring more top-tier AI and software engineering talent. Key focus areas include AI-driven business intelligence, ERP modernization, supply chain optimization, and CRM enhancement.
Strategy AI Platform Integrates AI Across All Apps
Strategy, formerly MicroStrategy, offers a new AI platform that integrates context-aware AI across all business applications. It provides a single governed layer for consistent intelligence, avoiding vendor lock-in and rebuilding. The platform uses a unified semantic layer to ensure every model speaks the enterprise's truth with consistent metrics and definitions. It also features enterprise-grade data security with encrypted transmission and a zero-retention policy. With over 200 native connectors, Strategy AI gives AI complete visibility into every system for accurate, end-to-end intelligence.
Chinese Tech Stocks and Chip IPOs Attract Investors
Beeneet Kothari, founder of Tekne Capital Management, believes the best tech investment opportunities are currently overseas, especially in China. He highlights China's advantage in AI and predicts a coming boom in chip IPOs there. Kothari also names Samsung as one of his favorite stocks. This suggests a shift in investor focus from the US to international markets for tech growth and innovation.
AI Reshapes Finance From Trading to New Startups
Artificial Intelligence is rapidly changing the finance industry, from trading desks to new startups. AI offers cheaper and faster solutions for tasks like paperwork and optimizes algorithmic trading. Platforms like Edge Hound help investors gain an edge with real-time data analysis and risk management. Fintech startups use AI for innovations in credit scoring, fraud prevention, and customer onboarding, attracting more early-stage investments. Regulators are increasing oversight to ensure transparency and prevent algorithmic bias as AI becomes more integrated into financial systems.
Tech Group Helps Democrats Catch Up on AI Campaigns
Tech for Campaigns, a Democratic-aligned group, is launching a new "AI Campaign Lab" to help Democrats use artificial intelligence effectively in elections. The group worries Democrats are falling behind in adapting to AI's impact on campaigning. The lab will partner with other groups in key races to fund experiments on different uses of AI, like shaping how candidates appear in AI chatbot results. These experiments, costing between $50,000 and $150,000, will share results widely to help campaigns replicate successful tactics. The goal is to ensure Democrats can leverage AI for personalized communication and platform engagement.
China's Minimax M2.5 Model Outperforms Rivals
Minimax, a leading Chinese AI company, released its new M2.5 large language model, which outperforms top systems from OpenAI, Anthropic, and Google on several benchmarks. M2.5 achieved 80.2% on the SWE-Bench Verified coding benchmark, beating GPT-5.2 (80.0%) and Gemini 3 Pro (78%). It also showed strong performance across other productivity and agentic evaluation suites. Minimax claims M2.5 is 37% faster for complex tasks and offers enterprise-ready throughput at a cost of just $1 per hour. This demonstrates China's rapid progress in AI development, closing the gap with US frontier models.
Sources
- Google Gemini 3 Deep Think Scores 3455 On Codeforces, Is Now Better Than All But 7 Human Programmers
- Google Releases Gemini 3 Deep Think, Tops ARC-AGI 2 Benchmark With 84.6%
- Gemini 3 Deep Think: Advancing science, research and engineering
- Bitcoin, Big Tech and the AI Trade Rollercoaster
- For stock market, AI turns from lifting all boats to sinking ships
- BT Business to offer AI training to all employees as it announces new AI apprentice and graduate roles
- Freight Brokers Are the Latest AI Victims. These Other Stocks Look Safe.
- Aeternum Tech LTD Expands International Platform Development Services Focused on Applied AI and Enterprise Execution Systems
- Strategy One
- The Allure of Chinese Tech Stocks—and a Coming Boom in Chip IPOs
- How AI Is Transforming Finance: From Trading Desks to Startup Disruption
- A tech group is launching a new effort to keep Democrats from falling behind on AI
- China’s Minimax Releases M2.5, Beats Gemini 3 Pro And GPT-5.2 On SWE-Bench
Comments
Please log in to post a comment.