• AI KATANA
  • Posts
  • Baidu launches two new AI models as industry competition heats up

Baidu launches two new AI models as industry competition heats up

Also: Y Combinator startups are fastest growing, most profitable in fund history because of AI

Author

AI KATANA
March 16, 2025

Hello!

The AI industry is evolving at a rapid pace, with major players like Baidu launching new models to compete in an increasingly crowded market. At the same time, the rise of AI-powered assistants is streamlining workflows for professionals, while concerns grow over AI’s influence on dating apps and personal relationships. Meanwhile, Y Combinator’s latest startup cohort is experiencing record growth, largely due to AI automation, and researchers remain skeptical about whether today’s models can truly achieve artificial general intelligence (AGI). On the tech front, Google’s experimental AI Mode is redefining search experiences, though its impact on content creators is still uncertain. In venture capital, Nvidia continues its aggressive expansion into AI startups, solidifying its dominance in the sector. From breakthroughs to ethical debates, today’s AI landscape is a mix of innovation, disruption, and unanswered questions.

Sliced just for you:

  • 🚀 Baidu launches two new AI models as industry competition heats up

  • 🤖 How to build your own AI assistant

  • ❤️ AI in dating apps ‘a threat to authentic intimacy’

  • 📈 Y Combinator startups are fastest growing, most profitable in fund history because of AI

  • 🧠 AI scientists are skeptical that modern models will lead to AGI

Baidu has introduced two new AI models—ERNIE 4.5 and X1—amid intensifying industry competition. ERNIE 4.5 is a multimodal model with advanced language, logic, and memory capabilities, while X1 is designed for reasoning and deep thinking, reportedly rivaling DeepSeek R1 at half the cost. Baidu highlights X1’s ability to autonomously use tools and its improved understanding, planning, and adaptability. Despite launching one of China’s earliest ChatGPT-style chatbots, Baidu has struggled to gain widespread adoption against leading U.S. AI models. The announcement underscores China’s aggressive push in AI innovation, with companies striving to produce cost-effective yet high-performing alternatives to U.S. models.

For frequent users of generative AI, the repetitive task of uploading background files and prompts can become inefficient, prompting many platforms to offer custom AI assistants. These assistants, such as ChatGPT’s “custom GPT,” Claude’s “Project,” and Google Gemini’s “Gem,” store essential instructions and files, streamlining workflows without requiring coding skills. Users can create assistants tailored to writing, marketing, troubleshooting, project management, and strategic coaching, improving productivity and decision-making. Setting up a custom assistant involves selecting a platform, experimenting with initial prompts, refining instructions, and supplying relevant files. Iterative feedback and adjustments help enhance accuracy, ultimately reducing the time spent fine-tuning responses in one-off AI interactions. By leveraging these assistants, professionals can optimize their workflow, focusing on core tasks while letting AI handle routine processes.

The increasing integration of AI into dating apps is raising concerns among academics and users about its impact on authentic intimacy. Companies like Match Group, which owns Tinder and Hinge, are investing in AI-driven features to refine user profiles, suggest date ideas, and improve conversations. While developers claim these tools enhance user expression and confidence, critics worry they may distort self-presentation, making interactions less genuine and potentially harming self-esteem. Researchers argue that AI-generated profiles and messages could lead to unrealistic expectations and further complicate an already challenging digital dating landscape. Some users report increased detachment and anxiety in online dating, with many taking breaks due to emotional fatigue. With declining user engagement on major dating platforms, there is growing interest in alternative methods, such as in-person dating events, signaling a potential shift in how people seek meaningful connections.

Y Combinator’s latest batch of startups is experiencing unprecedented growth, with AI playing a crucial role in accelerating development and profitability. According to CEO Garry Tan, AI is responsible for writing up to 95% of the code for a quarter of these companies, allowing them to scale rapidly with smaller teams and less capital. The entire batch of startups has been growing at an average of 10% per week over the past nine months, a rate never seen before in the accelerator’s history. With AI automating coding and other repetitive tasks, startups are reaching multi-million dollar revenues with lean teams of fewer than ten people. This shift comes amid a broader Silicon Valley transition from hypergrowth to a focus on profitability, as major tech firms like Google and Meta undergo layoffs. The rapid commercialization of AI-driven solutions is also attracting investor confidence, as many YC-backed companies are already proving real-world demand for their products. With 80% of the current YC startups being AI-focused, this trend signals a fundamental shift in early-stage entrepreneurship.

A survey of AI researchers reveals widespread skepticism about the likelihood of current AI models leading to artificial general intelligence (AGI). Despite massive investments from tech giants, 76% of respondents believe scaling existing approaches is unlikely to achieve AGI, as recent models show diminishing improvements. While companies continue to push AI expansion, many experts argue that the technology lacks fundamental understanding and problem-solving capabilities beyond narrow tasks. Researchers also highlight a disconnect between AI’s marketed capabilities and its actual performance, as systems still make significant errors despite claims of human-level proficiency. Meanwhile, companies like OpenAI and Google DeepMind define AGI differently, with some linking it to cognitive benchmarks and others tying it to profitability. This ongoing debate underscores the uncertainty surrounding AGI’s feasibility and the future direction of AI research.

🛠️ AI tools updates

Google’s new AI Mode in Search is a game-changer, blending AI Overview and Gemini chatbot functionalities to create a powerful, search-specific tool. Unlike previous AI-powered search features, AI Mode generates comprehensive, detailed responses without directly quoting sources, pulling insights from Google’s vast index. Users can ask multi-stage questions and refine their queries conversationally, making research more efficient. While AI Mode is still in experimental stages within Search Labs and requires a Google One AI Premium subscription for early access, it offers impressive results, consolidating information from multiple sources into structured, actionable summaries. The experience is streamlined, bypassing unnecessary conversational fluff while retaining interactivity, making it a valuable tool for quick, deep research. However, concerns remain about its impact on content creators, as it reduces reliance on traditional website visits. Despite this, for users needing fast, well-organized answers—whether for sports strategy, professional research, or general knowledge—AI Mode significantly enhances Google Search’s functionality, making it a preferred tool for those willing to embrace AI-driven search.

💵 Venture Capital updates

Nvidia has aggressively expanded its AI investment portfolio, solidifying its dominance in the industry by funding a range of high-profile startups. The company participated in 49 AI startup funding rounds in 2024, up from 34 in 2023, with significant contributions to companies advancing generative AI, autonomous systems, and AI-driven infrastructure. Major investments include OpenAI ($100M in a $6.6B round), Elon Musk’s xAI ($6B), Inflection AI ($1.3B), Wayve ($1.05B), and Scale AI ($1B). Nvidia has also backed startups in AI cloud computing (CoreWeave, Lambda), robotics (Figure AI), enterprise AI (Cohere, Perplexity), and AI-powered data centers (Crusoe). These investments extend beyond financial backing, as Nvidia seeks to accelerate AI innovation by ensuring startups have access to its powerful GPUs. With an eye on strategic market expansion, Nvidia’s venture activity underscores its ambition to shape the future of AI infrastructure and applications.

🫡 Meme of the day

⭐️ Generative AI image of the day