🚀 The first glimpse of GPT-5

Hey folks, 🚀 

Google is testing a machine learning model in the U.S. to estimate peoples ages, aiming to identify individuals under 18 and provide them with age-appropriate experiences across its platforms, including YouTube. The model analyzes data such as site visits, YouTube watch history, and account longevity to determine your age and allow you to verify if over 18. Big Brother isn’t just listening and watching, it’s acting on your content experiences too. Let’s dive in..

_{Read Time: 5 minutes} 

🧠 OpenAI lays out plans for GPT-5

 The Buzz: OpenAI is gearing up to drop GPT-4.5 soon, with GPT-5 not far behind, according to the wisdom of Sam Altman. He’s promised the new versions will try to untangle the current product web and make AI a bit more user-friendly. The new GPT-4.5, also known as Orion, will be the last of its kind before OpenAI moves to unified models, and it's all about making AI magic happen without the headaches.  

• GPT-4.5, or Orion, will be released within weeks and is set to be the last of the non-chain-of-thought models before the big shift to more unified systems. 

• GPT-5 will look more like an all-in-one interface that can handle whatever you throw at it, featuring voice, search, deep research, and more, all in one unified package.

• Users will get different levels of access based on their subscription plan.  

• OpenAI is focused on simplifying its product offerings to improve user experience with less model selection.  

💡 Takeaway: Knowing GPT-4.5 and GPT-5 are on the horizon makes it clear that OpenAI is working hard to make AI smarter and easier to use. For anyone dabbling in tech or relying on AI, these upgrades promise to change the way we interact with technology, which is exciting for both the casual user and the die-hard fan.

Together with The AI Report

There’s a reason 400,000 professionals read this daily.

Join The AI Report, trusted by 400,000+ professionals at Google, Microsoft, and OpenAI. Get daily insights, tools, and strategies to master practical AI skills that drive results.

Sign up now for free and work smarter, not harder.

🎥 ByteDance's Goku models create realistic product videos.

🐝 The Buzz: ByteDance has launched its Goku AI models, which can whip up realistic videos showing folks interacting with products—all without a single human actor in sight. The models were trained on a whopping 160 million image-text pairs and 36 million video-text pairs to crank out high-quality advertising clips. Goku+ aims to cut production costs by up to 99 percent compared to traditional methods.  

• Goku uses a cutting-edge transformer architecture with 2 to 8 billion parameters for creating images and videos from text prompts.  

• A specialized version, Goku+, focuses on realistic advertising featuring human interactions and gestures.  

• Benchmarks show Goku outperforms its competitors in quality, with the Goku-T2V scoring an impressive 84.85 on the VBench.  

💡 Takeaway: The arrival of Goku models could shake up the advertising world by making video production far cheaper and easier, challenging traditional creators. As companies hunt for fresh and budget-friendly video content, ByteDance’s offerings could become the go-to option for a savvy marketing strategy.

📰 Like newsletters? Here are more our readers enjoy.

💻️ How to add images onto videos

• Go to Pika labs and sign in

• Select “Get started with Pikaddition”

• Upload up to 5 secs video or use a sample

• Upload an image you want in the video

• Type your prompt: Describe what you want to do with the photo

• Click on the AI generate icon

🐝 AI buzz bits

⚡ Perplexity's new Sonar model, built on Llama 3.3 70B, delivers blazing fast answers at 1200 tokens per second, significantly enhancing factuality and readability, and outperforming competitors like GPT-4o and Claude 3.5 in user satisfaction and speed.

🦗 MIT has developed innovative tiny robotic insect drones that can now fly 100 times longer than their predecessors, making them potential for efficient and environmentally-friendly mechanical pollination of crops.

💻 Discover four open-source alternatives to OpenAI’s costly Deep Research AI Agent, providing powerful AI-driven research capabilities without the premium price tag, enabling flexibility, customization, and efficient information processing for researchers.

🏎️ Perplexity's bid to sponsor Red Bull Racing was thwarted by Oracle, its rival and the team's major sponsor, amid larger tech maneuvers involving TikTok partnerships and competition in the AI sector.

🖼️ Generative Art

A crisply focused close-up photograph of a a jellyfish floating in a gloomy forest creating a hyper-realistic, high-resolution, black & white, high-detail scene with cinematic lighting, soft shadows, depth of field, ominous, creepy, no contrast, clean sharp focus, and no graininess. The image was shot with a Canon EOS R5 at F2, ISO 300, and 86mm.

AI productivity tools
🔥 Take your workflow to the next level.

1. Sendbird - A powerful omnichannel AI customer service platform that delivers instant support across all digital channels 24/7.

2. FanHero - An all-in-one AI platform for creators to build communities, engage audiences, and monetize content effortlessly.

3. SWE Lens - Empower your hiring process with deep insights into junior software engineers' skills through GitHub project analyses.

4. Trainn - AI-driven no-code platform for creating customer training content, including videos, documentation, and knowledge bases.

For a full list of 1000+ AI tools, visit our Supertool Directory

🤘 Valentine’s Day Truth