- Human In The Loop
- Posts
- ChatGPT gets memory š§ , CapCutās viral age filter šµ, Copilot now runs Claude 4 š¤
ChatGPT gets memory š§ , CapCutās viral age filter šµ, Copilot now runs Claude 4 š¤
š” Insights
Coding
Windsurf is facing scrutiny as Anthropic limits its model access, spotlighting its struggles with high credit use and inconsistent performance, particularly for large-scale enterprise projects. Enter Qodo, GitHub Copilot, and Cursorāthree strong competitors stepping up with innovative solutions. Qodo leads for enterprises, boasting a Retrieval-Augmented Generation engine and strong multi-repo capabilities. This pivot signals a market shift toward more reliable, scalable AI coding tools, setting a new standard for productivity in software development.
At the TechCrunch Sessions: AI, Cohere's Solutions Architects, Yann Stoneman and Betsy Groves, highlighted the deployment of secure generative AI models tailored for regulated sectors like healthcare and finance, emphasizing data privacy by abstaining from external cloud reliance. This approach strategically positions Cohere as a privacy vanguard, appealing to industries prioritizing compliance and data sovereignty. It underscores a paradigm shift toward private AI systems, signaling to competitors the escalating demand for secure, in-house AI solutions.
Text
Perplexity AI is redefining its role by pivoting it from a chatbot to a search engine alternative, setting itself apart with features like source-backed summaries and autocomplete suggestions. While ChatGPT shines in nuanced tasks and conversational fluency, Perplexity bridges the gap between search engines and chatbots by focusing on fact verification and rapid, AI-generated summaries. This evolution not only challenges traditional search models but also forces competitors to rethink their strategy in the ever-evolving AI landscape.
Apple's integration of ChatGPT into its Image Playground elevates the company's AI prowess, blending generative AI with visual intelligence across iOS devices. This strategic shift underscores Apple's commitment to enhancing user experiences through advanced AI, setting a benchmark in the industry's AI arms race. Unlike the playful use of ChatGPT as a personal stylist, this move highlights Apple's vision for AI to foster more dynamic interactions, hinting at future capabilities that could redefine digital creativity and personalization.
Google's NotebookLM and Perplexity AI have been successfully integrated, creating a potent duo for streamlining research workflows. While NotebookLM provides robust organizational features like Mind Maps and Study Guides, Perplexity excels in fetching real-time, authoritative sources. This strategic pairing highlights a smart evolution in AI collaboration, where the synergy between tools transcends individual capabilities. For users, this means a more efficient, reliable research process, while the tech industry must now consider similar synergistic integrations as a pathway forward.
š ļø Product Updates
Video
CapCut's latest AI age filter, a viral sensation on TikTok, exemplifies the strategic use of AI-driven content creation in the social media landscape. This feature allows users to transform their selfies into aged versions using a template-driven approach, highlighting CapCut's innovative integration of artificial intelligence. As video editing apps like CapCut continue to enhance their offerings with cutting-edge tools, the competitive landscape shifts, pushing platforms towards more immersive and engaging user experiences. This positions CapCut and TikTok as front-runners in leveraging AI for creative expression.
Text
Google Sheets has launched a new feature allowing its AI assistant, Gemini, to create dynamic, fully editable charts that automatically update with changes in spreadsheet data. This shift from static images to interactive displays enhances user engagement and exemplifies Google's strategy to embed AI deeply into its products. However, the feature's limitations with large data sets and certain chart types highlight challenges in complexity. As the rollout progresses, users can expect increased productivity, setting Google apart in the AI-driven productivity tools market.
Perplexity's Comet browser is reportedly ramping up with new features, including a screenshot-taking function and planned Outlook integration for Windows. By allowing users to view visual snapshots and verify the AI agent's web navigation, Comet addresses challenges in handling UI-heavy or unstructured content. This positions Perplexity at the forefront of AI-driven automation, enhancing its cross-platform reach amid the competitive AI browser market. As Comet expands to Windows and Android, it could redefine enterprise productivity tools.
OpenAI has rolled out a transformative update for ChatGPT, offering users extensive customization options and enhanced features like "Customize ChatGPT" and "Memory," which personalize interactions. This strategic move indicates OpenAI's commitment to user-centric innovation, addressing privacy concerns while expanding functionality. The implications are significant: as users gain more control over their AI interactions, competitors must rapidly adapt to meet rising expectations in privacy and personalization, shaping the future landscape of generative AI.
Google has launched "Portraits," a new AI feature in Google Labs, initially in the US. This conversational AI model combines expert avatars with Geminiās reasoning to answer complex queries, starting with author Kim Scott. By attempting to merge authenticity with AI, Google is positioning itself as a leader in AI mentorship. However, concerns linger about AI's ability to responsibly handle complex human interactions, pointing to potential ethical challenges and opportunities in AI-driven guidance.
Anthropic's release of Model Context Protocol (MCP) servers significantly enhances Claude AI's capability to seamlessly integrate with applications like Gmail, Google Drive, and Spotify, offering direct, natural language interaction. This strategic move bypasses complex scripting, providing a user-friendly approach to AI-driven automation and contextual understanding. By simplifying tasks like email management and file organization, this development positions Anthropic as a front-runner in the AI integration space, potentially setting new standards for workflow automation across industries.
OpenAI plans to upgrade its $200 Pro subscription by replacing the o1 Pro model with the new o3 Pro model, designed to tackle complex tasks more efficiently. This move reflects a strategic push to enhance ChatGPT's capabilities, combining the o3 model's superior reasoning with the Pro tier's computational power. For users, this means a more powerful AI tool for advanced math, coding, and scientific research, while competitors must keep pace in the race for robust AI development.
Coding
Tabnine has unveiled an AI software development platform tailored for aerospace and defense, overcoming key drawbacks of traditional AI code assistants. By offering sovereign, air-gapped deployment, deep integration with enterprise contexts, and extensive support for mission-critical languages, Tabnine addresses the industry's need for airtight security, traceability, and compliance. Positioned ahead of cloud-dependent competitors like GitHub Copilot Enterprise, Tabnine's solution signals a shift towards secure, auditable AI development in high-stakes sectors, setting a new benchmark for regulated environments.
GitHub just enhanced Copilot by integrating Anthropic's Claude 4 models, offering developers more advanced coding assistance for both code refactoring and review. This strategic upgrade strengthens GitHub's position as a leader in AI-driven code tools, showcasing a commitment to evolving beyond basic suggestions into nuanced, context-aware guidance. For users, this means higher-quality code maintenance and faster development cycles, while competitors face heightened pressure to innovate in the race for AI-enhanced coding solutions.
Audio
ElevenLabs has unveiled its v3 text-to-speech model, boasting advanced expressiveness and the ability to convey emotions across 70+ languages, leaping past its predecessorās 29. This major upgrade elevates human-machine interaction by incorporating subtle cues like sighs and laughter, a feature that not all find convincing yet. Currently in public alpha with an enticing 80% discount, v3 sets a new standard in TTS, challenging competitors like Hume AI and Google to rethink emotional nuance and linguistic scope.
ā Reviews
HeyGen faces intensified competition as six AI video platforms are spotlighted in a recent review, each offering unique advantages over HeyGen's current offerings. Notably, Synthesia stands out for its corporate training capabilities with over 230 avatars, while Pictory efficiently transforms text into social videos. This competitive landscape suggests that while HeyGen is popular, users increasingly seek more flexibility and features, prompting a call to action for HeyGen to innovate and possibly rethink pricing strategies to maintain its position.
š How-To
A new tutorial expertly demonstrates an advanced research workflow by integrating SerpAPI with Google's Gemini-1.5-Flash model within a Google Colab notebook, featuring the AdvancedSerpAPI Python class. This strategic move enhances AI-driven research by merging powerful search capabilities with smart analytics ā a significant leap ahead in simplifying research tasks. Users benefit from streamlined, modular solutions for web-based data collection and insight synthesis, setting a new standard for research efficiency in tech-driven fields.
Microsoft has unveiled nine new tips for mastering its innovative Microsoft 365 Copilot, an AI feature in tools like Word, Excel, and PowerPoint. By simplifying onboarding and offering features such as AI-powered summaries and reusable prompts, Copilot is designed to streamline workflows without requiring technical know-how. This strategic enhancement underscores Microsoftās commitment to leading in AI-driven productivity, offering users an edge in efficiency and collaboration, likely setting a high bar for competitors in the AI productivity space.
š§Ŗ Use Cases
Chinese AI start-up DeepSeek is recruiting interns to label medical data, aiming to enhance its AI capabilities in hospitals, particularly for advanced auxiliary diagnosis tools. This move underscores DeepSeek's proactive strategy of bolstering its low-cost open-source models, which are already deployed in 300 Chinese hospitals. As the company intensifies its focus on healthcare, this initiative could further solidify DeepSeekās standing in the AI sector, challenging global giants by offering specialized, sector-focused solutions without disclosing a release date for its R2 reasoning model.
Reply