google, deepseek and microsoft Updates

The artificial intelligence landscape is experiencing rapid shifts, with Google's Gemini making significant inroads against ChatGPT. ChatGPT's share of generative AI traffic recently fell below 70% for the first time, while Gemini surged to 18.2%, gaining 4.5 percentage points in just one month. Other players like Grok, Perplexity, and Claude also show growth, though DeepSeek and Microsoft Copilot have seen less momentum, indicating a highly competitive market focused on unique features and integration. This competition extends to device integration, as Samsung's Bixby assistant is enhancing its capabilities by integrating with Perplexity AI. This move suggests a strong software focus for future devices like the Galaxy S26. Soon, Galaxy phone users may choose between Google's Gemini, Samsung's own AI, and Perplexity for various tasks, mirroring Apple's reported plans for Siri and ChatGPT integration. Meanwhile, physical AI is transforming industries such as mining, energy, and agriculture by embedding AI directly into machinery for improved predictability and operational efficiency. However, the rapid advancement of AI also brings significant concerns and regulatory challenges. Geoffrey Hinton, often called the "Godfather of AI" and a 2024 Nobel Prize winner, warns that artificial intelligence could potentially destroy humanity, recalling past skepticism about neural networks. In Colorado, lawmakers are grappling with regulating high-risk AI systems. Senate Bill 205, effective in 2026, aims to protect consumers from algorithmic discrimination in areas like jobs and loans, requiring developers to exercise reasonable care and conduct impact assessments, with potential $20,000 fines per violation. Legal scrutiny is also increasing, as author Elizabeth Lyon is suing Adobe, alleging the company used pirated versions of her books to train its SlimLM language model. This case is part of a growing trend of legal action against tech giants, including Apple and Salesforce, for using copyrighted material without permission to train AI models. Separately, the reliability of AI detection tools in education faces criticism, with instances like 17-year-old Ailsa Ostovitz being wrongly accused of using AI on assignments, despite researchers confirming the unreliability of popular detectors like Turnitin and GPTZero. Internationally, Israel aims to become a global leader in AI. Prime Minister Benjamin Netanyahu and Elon Musk discussed promoting AI technologies and advancing autonomous vehicle legislation in Israel, with Musk planning a visit in March for a Smart Transportation Conference. Globally, the United States currently leads the AI race in research, development, and investment, but China is rapidly closing the gap, boasting more AI engineers and significant company investment. The year 2025 marked a turning point for AI, prompting public and expert examination of its overall impact.

Key Takeaways

• ChatGPT's generative AI traffic share dropped below 70%, while Google's Gemini surged to 18.2%, gaining 4.5 percentage points in one month.
• Samsung's Bixby assistant is integrating Perplexity AI, allowing Galaxy users to choose between Google's Gemini, Samsung's own AI, and Perplexity for tasks.
• Apple is reportedly planning to integrate ChatGPT with Siri, similar to Samsung's multi-AI approach.
• Geoffrey Hinton, "Godfather of AI" and 2024 Nobel Prize winner, warns that AI could potentially destroy humanity.
• Colorado's Senate Bill 205, effective 2026, will regulate high-risk AI systems to prevent algorithmic discrimination, imposing $20,000 fines per violation.
• Adobe faces a lawsuit from author Elizabeth Lyon for allegedly training its SlimLM model on pirated books from the Books3 dataset.
• Other tech giants like Apple and Salesforce also face legal action for using copyrighted material to train AI.
• AI detection tools used in education, such as Turnitin and GPTZero, are often unreliable and have led to false accusations against students.
• Israel's Prime Minister Benjamin Netanyahu and Elon Musk discussed promoting AI and autonomous vehicle legislation in Israel, with Musk planning a March visit.
• The United States leads the global AI race in research and investment, but China is quickly catching up with more AI engineers and heavy company investment.

Geoffrey Hinton warns AI could destroy humanity

Geoffrey Hinton, known as the "Godfather of AI," won the 2024 Nobel Prize in Physics for his work on artificial neural networks. He also received the 2018 Turing Award with collaborators Yoshua Bengio and Yann LeCun. Hinton worked at Google's AI team from 2013 to 2023 and is now a professor emeritus at the University of Toronto. Despite his contributions, he warns that artificial intelligence could potentially destroy humanity. He recalls how many experts once thought neural networks were not a good approach for AI.

Colorado struggles to regulate AI amid job market concerns

Colorado lawmakers face a deadlock over regulating high-risk AI systems, even as residents like Taahaa Dawe struggle in the job market. Senate Bill 205, signed by Governor Jared Polis, aims to protect consumers from algorithmic discrimination in important decisions like jobs and loans. This law, set for 2026, requires AI developers to use reasonable care, provide disclosures, and conduct impact assessments. Companies could face $20,000 fines per violation for discrimination. Experts like Robin Burke from the University of Colorado highlight how biased data can lead to unfair outcomes, citing a lawsuit against Workday for alleged job discrimination.

Bill Gurley discusses AI and career success on Tim Ferriss Show

Venture capitalist Bill Gurley appeared on The Tim Ferriss Show on December 17, 2025, to discuss his new book, "Running Down a Dream." Gurley shared lessons on how successful people like Jerry Seinfeld and Bob Dylan mastered their crafts through obsession. He also offered his views on the current AI landscape and how individuals can stay relevant in a changing economy. Gurley's book explores how Jerry Seinfeld found inspiration to pursue comedy after reading "The Last Laugh" by Phil Berger. The conversation also touched on the idea of an AI bubble.

Netanyahu and Musk discuss AI Israel visit planned

Prime Minister Benjamin Netanyahu spoke with US billionaire Elon Musk in a conference call from Florida. They discussed promoting AI technologies and advancing legislation for autonomous vehicles in Israel. Transportation Minister Miri Regev and National AI Directorate head Erez Askal also joined the call. Musk plans to visit Israel in March to attend a Smart Transportation Conference. Netanyahu stated his goal is to make Israel a global leader in AI, similar to its success in cyber technology.

Netanyahu and Musk discuss AI future for Israel

Prime Minister Benjamin Netanyahu and Elon Musk held a conference call to discuss the future of artificial intelligence in Israel. They also talked about working together on autonomous vehicle technology. The Prime Minister's Office confirmed the call. This discussion happened days after Musk faced criticism for a post on X about Jews and whiteness.

2025 marked a turning point for AI

The year 2025 proved to be a significant period for artificial intelligence. During this time, the public and experts alike closely examined AI's overall impact and how it was being received. This "vibe check" likely involved discussions about its benefits, challenges, and future direction.

Gemini challenges ChatGPT's AI market dominance

ChatGPT's share of generative AI traffic has dropped below 70% for the first time, signaling a major shift in the market. Gemini, Google's AI assistant, has rapidly gained ground, surging to 18.2% of the market share. A year ago, ChatGPT held 87.2%, but Gemini's recent strategic product releases have helped it gain 4.5 percentage points in just one month. Other AI players like Grok, Perplexity, and Claude also show growth, while DeepSeek and Microsoft Copilot have seen less momentum. This indicates the AI market is becoming more competitive, focusing on unique features and integration.

Samsung Bixby gets smarter with Perplexity AI integration

Samsung's Bixby assistant is becoming smarter by integrating with Perplexity AI, a move that hints at the Galaxy S26's focus on software. Some users already report Bixby using Perplexity for tasks like weather forecasts. Soon, Galaxy phone users can choose between Google's Gemini, Samsung's own AI, and Perplexity for different tasks. Bixby will handle simple commands, while Perplexity will manage complex searches, similar to Apple's plan with Siri and ChatGPT. This integration could bring Bixby back as Samsung's main assistant. The second One UI 8.5 beta update, focusing on bug fixes, is also out for Galaxy S25 models in several countries.

Physical AI transforms mining energy and agriculture

Physical artificial intelligence is changing industries like mining, energy, and agriculture, which were once slow to adopt new technology. This AI is now built directly into machinery, acting as a critical part of operations in harsh environments. These systems focus on predictability, allowing machines to make quick decisions without needing a distant server. In mining, AI improves planning and productivity while humans keep control. For agriculture, AI helps farmers manage complex operations by handling perception and small adjustments. Energy companies like Saudi Aramco use high-performance computing to process data, detecting equipment issues early and reducing downtime.

US leads AI race but China is catching up

The United States currently leads the global artificial intelligence race, but China is quickly closing the gap. The US has an advantage in AI research, development, and investment, with more top researchers and venture capital funding. However, China boasts more AI engineers and its companies are investing heavily in the field. China's large market for AI products also drives its innovation. To maintain its lead, the US must continue investing in AI research and development, support startups, and ensure a strong pool of AI talent.

Law professor uses AI in contracts course

Professor Gregory Duhl at Mitchell Hamline School of Law will teach a new contracts course this spring that includes generative artificial intelligence. Students in the class will learn to build bots by teaching them legal doctrine. The course will also use Socratic dialogue bots throughout the semester. A key part of the class involves learning how to detect "hallucinations" or errors in judicial opinions generated by AI.

Adobe faces lawsuit for training AI on pirated books

Adobe faces a lawsuit from author Elizabeth Lyon, who claims the company used pirated versions of her books to train its SlimLM language model. The lawsuit states that SlimLM was trained using the SlimPajama-627B dataset, which includes content from the Books3 dataset containing over 190,000 books. This case is part of a growing trend of legal action against tech giants like Apple and Salesforce for using copyrighted material without permission to train AI. Marketers are advised to understand their AI tools' data sources, audit content workflows, and include AI indemnity in contracts to avoid legal risks.

AI detection tools wrongly accuse students

Teachers are increasingly using AI detection software, but these tools sometimes wrongly accuse students of cheating. Ailsa Ostovitz, a 17-year-old student, was accused of using AI on three assignments, with one detector showing a 30.76% probability. Her teacher docked her grade, though later changed their mind after a meeting with Ostovitz's mother. Prince George's County Public Schools advises against relying on these tools due to their inaccuracies. Researchers like Mike Perkins confirm that popular AI detectors like Turnitin, GPTZero, and Copyleaks are often unreliable. Despite this, some large districts like Broward County Public Schools still spend significant money on them.

Sources

• The ‘Godfather of AI’ says we can’t afford to get it wrong
• Coloradans are feeling the effects of AI as the state’s legislative deadlock over regulation continues
• Tim Ferriss Show: w/ Bill Gurley on Navigating the AI Era (Transcript)
• Netanyhu talks to Elon Musk about AI, autonomous cars; billionaire says he’ll visit Israel in March
• Netanyahu, Elon Musk discuss future of artificial intelligence in Israel in conference call
• 2025 was the year AI got a vibe check
• ChatGPT’s Traffic Share Drops Below 70% For First Time, Gemini Surges to 18.2%
• Galaxy S26 will be all about software, not hardware: Bixby is already getting smarter with Perplexity AI
• From Mines to Oil Rigs, How AI Is Rewiring Mining, Energy and Agriculture
• The AI scorecard: How the US built a lead—and could lose it to China
• Mitchell Hamline Law Professor Integrates AI Into Contracts Course| Law.com
• Adobe sued over alleged AI training with pirated books
• Teachers are using software to see if students used AI. What happens when it's wrong?