google, openai and gemini Updates

The race in artificial intelligence continues to intensify, with Google's Gemini 3 Pro currently leading OpenAI's GPT 5.2 (High) on the Artificial Analysis Intelligence Index. Gemini 3 Pro matched GPT 5.2 (High) with an intelligence score of 73, but it proved faster, achieving a speed score of 119 compared to GPT 5.2's 90. Furthermore, Gemini 3 Pro offers a more cost-effective solution at $4.5 per million tokens, while GPT 5.2 costs $4.8 per million tokens. Despite GPT 5.2 (High) showing improvement over its predecessor, GPT 5.1, it still trails Gemini 3 Pro overall. However, not all AI performance is about benchmark scores. Lovable CEO Anton Osika highlighted a surprising limitation of GPT 5.2's "Instant" version, demonstrating its struggle with simple tasks like accurately counting letters in a word. This issue stems from how large language models break words into "tokens" rather than individual letters, leading some experts, like former OpenAI Chief Scientist Ilya Sutskever, to suggest a phenomenon called "benchmark-maxing." Meanwhile, former Google CEO Eric Schmidt expressed amazement at AI's rapid improvement in coding abilities, predicting that artificial general intelligence (AGI) could arrive within three to five years, potentially by 2029. Schmidt believes AGI will match the intelligence of the best human experts and significantly automate business tasks such as billing and accounting. In response to the fierce competition, Google co-founder Sergey Brin returned to hands-on AI development in February 2023, a move Google CEO Sundar Pichai credits with significantly boosting the company's AI strategy and products, including Gemini. On the political front, President Donald Trump signed an executive order to establish a single federal approach for AI regulation, limiting states' power. This decision, supported by tech advisors like Peter Thiel and Elon Musk, faced opposition from some MAGA supporters who feared it would harm competition and consolidate control among large tech companies. Beyond the tech giants, AI is finding diverse applications and facing new challenges. Churches in the U.S. and abroad are leveraging AI for sermon ideas, visuals, and volunteer scheduling, helping them reach wider audiences despite declining attendance. Yet, concerns persist about AI-generated content lacking "heart" and perpetuating biased stereotypes in images. Businesses are also actively planning their AI integration for 2025, prioritizing data readiness, security against breaches, clear usage guidelines, and ethical considerations. Even the publishing world is feeling the impact, with Springer Nature facing criticism for an AI ethics guide containing questionable and fake citations. Despite these complexities, analysts believe advertising agencies can benefit from AI advancements, using tools like Google's Nano Banana and OpenAI's Sora 2 to manage complex media landscapes, while global fund managers eye Indian stocks as a stable investment against potential AI market volatility.

Key Takeaways

• Google's Gemini 3 Pro leads OpenAI's GPT 5.2 (High) on the Artificial Analysis Intelligence Index, tying in intelligence score (73) but surpassing it in speed (119 vs 90) and cost ($4.5 vs $4.8 per million tokens).
• OpenAI's GPT 5.2 (Instant) struggles with basic tasks like counting letters in a word, as demonstrated by Lovable CEO Anton Osika, raising concerns about "benchmark-maxing" where models perform well on tests but not real-world tasks.
• Former Google CEO Eric Schmidt predicts artificial general intelligence (AGI) will arrive within three to five years, potentially by 2029, and emphasizes AI's significant impact on automating business tasks and coding.
• Google co-founder Sergey Brin returned to active AI development in February 2023, a move credited by CEO Sundar Pichai as crucial for enhancing Google's AI strategy and products, including the Gemini model.
• President Donald Trump signed an executive order to establish a unified federal approach for AI regulation, limiting states' power, a decision supported by tech advisors but opposed by some MAGA supporters concerned about competition.
• Businesses are actively planning AI integration for 2025, focusing on data readiness, robust security measures, clear usage guidelines, and ethical considerations to avoid common implementation pitfalls.
• Churches are utilizing AI tools for various activities like sermon generation, visual content, and volunteer scheduling, though some leaders express reservations about the emotional depth of AI-generated content and potential for biased imagery.
• Advertising agencies are poised to benefit from AI advancements, with analysts suggesting major brands will increasingly rely on agencies to navigate complex media landscapes and leverage new AI tools like Google's Nano Banana and OpenAI's Sora 2.
• Global fund managers are considering Indian stocks as a strategic diversification option to mitigate potential risks associated with an AI market bubble, citing India's strong domestic growth and market independence.
• Publisher Springer Nature faced criticism for an AI ethics guide containing numerous questionable and fake citations, including references to non-existent journals, raising concerns about the accuracy of published AI-related content.

GPT 5.2 struggles with simple word counting says CEO Anton Osika

GPT 5.2 scored high on AI tests like ARC-AGI 2. However, Lovable CEO Anton Osika showed it struggles with simple tasks, like counting letters in a word. This happens because AI models like LLMs break words into "tokens" instead of single letters. The specific answer Osika shared came from GPT 5.2's "Instant" version, which is its least capable. Some experts, like former OpenAI Chief Scientist Ilya Sutskever, suggest this could be "benchmark-maxing" where models perform well on tests but not real-world tasks.

OpenAI GPT 5.2 ranks second to Google Gemini 3 Pro

OpenAI's new GPT 5.2 (High) model placed second on the Artificial Analysis Intelligence Index. Google's Gemini 3 Pro took the top spot, tying GPT 5.2 (High) with a score of 73 on the Intelligence Index. The index tests AI models on math, science, and coding skills. Gemini 3 Pro proved faster with a speed score of 119 compared to GPT 5.2's 90. Gemini 3 Pro was also cheaper at $4.5 per million tokens, while GPT 5.2 cost $4.8 per million tokens. GPT 5.2 (High) did rank second in coding and agentic indexes, showing it is an improvement over GPT 5.1 but still behind Gemini 3 Pro overall.

Churches use AI for Christmas programs and outreach

Churches in the U.S. and other countries are using AI tools for their Christmas activities. They use AI to create sermon ideas, generate visuals for services, schedule volunteers, and make social media content. This helps churches with limited staff and resources to reach more people, especially as attendance declines. Experts like Rev. Chris Hope say AI can be a helpful tool for ministry, but not the source of it. However, some, like Rev. Malcolm Young, find AI-generated content "flat" and without "heart." Concerns also exist about AI images showing biased stereotypes, like a white Jesus, and the need to protect in-person community.

Ex-Google CEO Eric Schmidt amazed by AI coding abilities

Former Google CEO Eric Schmidt expressed surprise at how fast AI is improving, especially its ability to write code by itself. He believes AI's biggest impact will be automating business tasks like billing and accounting, which currently cost billions. Schmidt predicts that artificial general intelligence, or AGI, will arrive within three to five years, possibly by 2029. AGI systems will be as smart as the best human experts in many fields. He warned that humans must oversee AI to keep our freedom and control.

Trump signs AI order despite MAGA opposition

President Donald Trump signed an executive order that limits states' power to regulate artificial intelligence. His tech advisers, including Peter Thiel and Elon Musk, supported this move to create a single federal approach for AI development. However, some of Trump's MAGA supporters strongly opposed the order. They worried it would hurt competition and let large tech companies control the AI industry. This event shows how Silicon Valley's influence on Trump's plans is growing and highlights disagreements within his political group.

Sergey Brin's return boosts Google's AI efforts

Google co-founder Sergey Brin returned to hands-on AI work in February 2023. Google CEO Sundar Pichai said Brin's involvement was key to improving the company's AI strategy and products, like the Gemini AI model. Brin, who had stepped back in 2019, came back due to strong competition in the AI field from companies like OpenAI and Microsoft. Pichai highlighted Brin's deep technical knowledge and unique perspective as a co-founder. Many see Brin's active role, including coding and reviewing AI products, as a major boost for Google in the intense AI race.

India stocks offer safe bet against AI market risks

Global fund managers are looking at Indian stocks to spread out their investments next year. They see India as a good way to protect against possible risks from an AI market bubble. Experts like Raj Singh from Principal AMC say India's market does not move closely with other markets. India has strong growth within its own country, helped by tax cuts, new labor laws, and supportive government policies. Christina Woon of Eastspring Investments and Jerry Goh of Aberdeen also agree that India offers good diversification and reasonable earnings expectations.

Businesses plan AI use balancing innovation and security

Businesses are creating plans to use AI, focusing on both new ideas and keeping systems safe in 2025. Companies must ensure their data is ready and secure, checking for quality and following privacy rules. Clear rules are needed for how AI is used to prevent problems, as many firms have AI tools but no clear guidelines. Security is also key, with measures against data leaks and attacks, as most breached organizations lack proper controls. It is important to train employees, manage costs, and make sure AI works well with existing systems. Leaders must also consider ethical issues and how to measure AI's success, avoiding common mistakes like treating AI as a simple plug-and-play tool.

Publisher faces criticism for fake citations in AI ethics book

Publisher Springer Nature is facing criticism because an AI ethics guide it published contains many questionable citations. These citations include references to journals that do not actually exist. This issue raises concerns about the quality and accuracy of information in the AI ethics book.

Ad agencies can benefit from AI changes say analysts

Analysts believe advertising agencies can use AI advancements to their benefit, despite initial worries about disruption. Major brands will rely more on agencies to manage complex media landscapes, especially with new AI tools like Google's Nano Banana and OpenAI's Sora 2. Agencies like Publicis and Omnicom are getting strong recommendations because they help clients wisely spend ad budgets across many platforms. Experts like Mark Giarelli from Morningstar say agencies understand consumer behavior data and can tailor messages effectively. Lower production costs from AI could also lead to more advertising and bigger investments from top brands. While some agencies face challenges, consolidation in the industry could offer new opportunities.

Sources

• Lovable CEO Anton Osika Highlights How Even GPT 5.2 Can’t Count The Number Of ‘R’s In A Word
• GPT 5.2 High Places Second Behind Gemini 3 Pro In Artificial Intelligence’s Intelligence Index
• It's beginning to look a lot like (AI) Christmas
• 'Holy crap. The end of me', says ex-Google CEO Eric Schmidt on AI doing coding jobs - The Times of India
• How Trump’s tech advisers overcame a MAGA rebellion over AI
• Did Google co-founder Sergey Brin’s return save Google in the AI race? Sundar Pichai explains | - The Times of India
• Global Funds View Indian Stocks as a Top Hedge Against AI Risks
• Crafting AI Enablement Plans: Balancing Innovation and Security in 2025
• Publisher under fire after ‘fake’ citations found in AI ethics guide
• Ad Agency Stocks Seen Turning AI Disruption to Their Advantage