Google Rolls Out Gemini 2.0 lineup🤯

Google DeepMind just released three new AI tools that are making waves. Think of them as a family - there's Gemini 2.0 Flash, which is ready to help developers build real-world apps. Then there's Gemini 2.0 Pro, the experimental one that's really good at coding and tough problems. And finally, there's Gemini 2.0 Flash-Lite, the practical one that won't break the bank. Right now, they can all handle different types of input (like text and images) and turn them into text, with more cool features coming soon.

Meanwhile, there's an exciting story of old friends starting something new. Remember ChatGPT? One of its creators, John Schulman, has joined forces with Mira Murati, who used to be a top leader at OpenAI. After a quick stop at another company called Anthropic, Schulman decided to team up with Murati on her new project. They're gathering some of the smartest people in AI from companies like OpenAI, Character AI, and DeepMind. With plans to raise $100 million, they're cooking up something that could change the AI world.

And speaking of changes, OpenAI has given itself a fresh new look! They've created a special logo called the Blossom, which looks like a perfect circle mixed with technology shapes. They've even made their own font called OpenAI Sans. But this isn't just about looking good - it's about showing how AI and humans can work together to create amazing things.

From new AI tools to fresh starts and makeovers, February has been full of surprises in the AI world. Let's dive in and explore these stories that are shaping our future with technology!

Google DeepMind Updates Gemini 2.0 Models

Google DeepMind has released updates to its Gemini 2.0 models, including Gemini 2.0 Flash, Gemini 2.0 Flash-Lite, and Gemini 2.0 Pro Experimental. Gemini 2.0 Flash is now generally available via the Gemini API in Google AI Studio and Vertex AI, allowing developers to build production applications. An experimental version of Gemini 2.0 Pro, optimized for coding performance and complex prompts, is also available in Google AI Studio and Vertex AI, as well as to Gemini Advanced users in the Gemini app. Gemini 2.0 Flash-Lite, a new cost-efficient model, is in public preview in Google AI Studio and Vertex AI. All of these models support multimodal input with text output, with more modalities planned for future releases.

OpenAI Co-Founder Schulman Joins Murati's New AI Startup

John Schulman, a co-founder of OpenAI and a key architect of ChatGPT, is joining Mira Murati's new, still unnamed startup after a brief five-month stint at Anthropic. Murati, who departed from her role as CTO at OpenAI in September 2024, is seeking over $100 million in funding for the venture, which focuses on groundbreaking AI innovations. Schulman had left OpenAI in August 2024 to focus on AI alignment and engage in more technical work. Murati has also recruited other former OpenAI employees, including Christian Gibson from the supercomputing team, Jonathan Lachman, former head of special projects, and Mario Saltarelli, a longtime IT manager. Other high-profile talent from companies like Character AI and Google's DeepMind have also joined Murati's startup.

OpenAI Branding: A Guide to Logos and Trademarks

OpenAI's brand guidelines detail how third parties should use OpenAI's trademarks and copyrighted materials, including the "OpenAI" name, logo, "ChatGPT," and "GPT" brands. The guidelines cover the proper use of the OpenAI wordmark (created with OpenAI Sans, with the 'O' as a perfect circle) and the Blossom logo, which represents the intersection of humanity and technology. The Blossom should be used primarily for OpenAI Research with prescribed spacing. The guidelines outline specific do's and don'ts for using both logos, emphasizing avoiding alterations, maintaining clear space, and not using the Blossom as primary branding outside of OpenAI Research. When using the logos, it should directly relate to OpenAI services, follow OpenAI’s style guides, and not misrepresent the relationship with OpenAI. The document also provides guidance on typography, brand partnerships (emphasizing balanced spacing and approval processes), language usage, content attribution, and press releases. OpenAI has also undergone a rebrand, introducing a new logo, typeface (OpenAI sans), and color palette to create a more organic and human identity. Permission to use OpenAI’s Marks is limited, requiring adherence to the guidelines and not featuring the Marks more prominently than one's own. For inquiries and permission requests.

Hand Picked Video

In this video, we'll look at Gemma 2 Model, Benchmarks, how to run locally and finally how it works after i integrated it in my app.

Top AI Products from this week

• Concierge AI - Concierge lets you connect to your SaaS apps (e.g. Gmail, Slack, Notion, etc) with just a few clicks. Select any of the latest and greatest models, including GPT, Claude, Grok, and yes, Deepseek. Then simply ask whatever you want.

• Lyzr Agent Studio - Lyzr Agent Studio empowers enterprises to build, test, and deploy AI agents with confidence. With built-in safeguards for accuracy and safety, it’s an all-in-one platform that takes agents from POC to production—streamlining workflows.

• Kompas AI - Kompas AI conducts multi-depth research across hundreds of web pages, saving you time by filtering out irrelevant data. It also streamlines long-form report creation and editing, allowing you to easily generate comprehensive insights.

• Jam AI - Never write repro steps again! Now you can write bug reports with AI in 1 click. Just record a bug and let Jam do the rest. AI writes the technical description and steps to reproduce from your video.

• Airtop Auth - Airtop is revolutionizing web automation with AI-powered cloud browsers that seamlessly handle authentication. Whether you're building AI agents, automation tools, or web scrapers, Airtop lets you control browsers in the cloud without getting stuck on logins.

• Digma Preemptive Observability - Digma is a Preemptive Observability product that identifies performance and scaling issues in pre-production environments and provides a code-level root cause and the severity of the issues, all done automatically and continuously.

This week in AI

• Vision Search Assistant - Vision Search Assistant enhances Vision-Language Models using web knowledge. It identifies objects, generates sub-questions, and analyzes web pages to answer questions about novel images and complex scenarios. It outperforms Perplexity.ai Pro and GPT-4o-Web.

• OpenAI Hardware - OpenAI's trademark application reveals potential AI-powered consumer hardware, including wearables, robots, and VR/AR headsets123.CEO Sam Altman aims to create AI-first hardware, possibly with Jony Ive.

• EU AI Act Phase 1 - As of February 2, 2025, the EU AI Act's first phase is in effect, banning AI systems posing "unacceptable risks"138.Companies must ensure AI literacy among employees12.Non-compliance can result in hefty fines2.The Act categorizes AI risks into four levels

• Google vs. DeepSeek - Google CEO Pichai claims Gemini models are more efficient than DeepSeek's, despite DeepSeek impacting Nvidia's market value. Google Cloud revenue grew 30% YoY, with plans to invest $75B in infrastructure in 2025.