Read Online | Sign Up | Advertise | | | Good morning, AI enthusiasts. We're still in the first month of 2025, but a new OpenAI reasoning model is now reported to be right around the corner. | The coming release of o3-mini, along with rumblings of 'PhD level SuperAgents' and a new human longevity research model, suggest OpenAI may be set for the next major move up the AI ladder. | Friendly reminder: Our newsletter is still slowly moving to a new email address. Add news@daily.therundown.ai to your contacts to ensure you keep receiving our emails. | | In today's AI rundown: | OpenAI readies 'o3-mini' model launch Altman to brief Washington on 'PhD level SuperAgents' Get automatic global news briefings with ChatGPT Runway releases 'Frames' image generation model 4 new AI tools & 4 job opportunities
|
|
| | | | OPENAI | |  | Image source: OpenAI |
| The Rundown: OpenAI CEO Sam Altman announced that the company has finalized its anticipated 'o3-mini' reasoning model for release in the coming weeks, with plans to launch API access and ChatGPT integration simultaneously. | The details: | o3-mini is the next iteration of OpenAI's reasoning models, following September's o1, which showed enhanced science, coding, and math abilities. The company plans to release the model simultaneously through its API and ChatGPT, a departure from previous staggered releases. Altman said that OpenAI is now focusing on o3 and o3-pro models, which he hinted will be available to users in the $200/m Pro tier. Altman also commented that o3-mini is "worse than o1 pro at most things, but FAST" when asked to compare it to the startup's current premium model.
| Why it matters: OpenAI's first model release of 2025 is arriving shortly, and given the excitement coming internally (more below) and the rate of improvement already happening, the AI launch cycle will likely be set for a massive acceleration in the coming year. |
|
| | TOGETHER WITH UNSTRUCTURED | | | The Rundown: Join Unstructured for a live session on how to fuel your vector search and training pipelines with clean, structured data — a key necessity for building great GenAI apps. | Join the live webinar on tomorrow and learn how to: | Efficiently convert documents like text, tables, images, and more into JSON Build scalable ETL pipelines with multiple data sources and destinations See a live code demo of actionable steps to unlock internal business data
| Click here to register and receive the recording, even if you can't attend live. |
|
| | OPENAI | |  | Image source: Reuters |
| The Rundown: OpenAI CEO Sam Altman is set to brief U.S. government officials on a major AI advancement at the end of January, with sources suggesting the company has developed sophisticated AI agents capable of tackling complex tasks at an expert level. | The details: | Altman has scheduled a closed-door presentation in Washington on Jan. 30th to showcase "PhD level" AI systems capable of complex problem-solving. The meeting was revealed in OpenAI's recent U.S. Economic Blueprint, which outlined initiatives necessary to usher in an era of "shared prosperity." Axios's report also revealed that OpenAI staff have been "jazzed and spooked by recent AI progress." OpenAI also developed GPT-4b micro, a model that engineers proteins for cellular reprogramming — with results 50x more effective than those of scientists.
| Why it matters: While consumers are getting o1 Pro and o3-mini, it looks like many bigger advances are happening behind closed doors at OpenAI. With talk of "PhD level SuperAgents" and a new longevity-specialized model showcasing new breakthroughs, AI's fast takeoff could happen right before our eyes. |
|
| | AI TRAINING | | | The Rundown: ChatGPT's new Tasks feature lets you receive personalized daily news summaries highlighting key global events and surprising developments, all curated to your interests. | Step-by-step: | Select "GPT-4o with Scheduled Tasks" in ChatGPT. Ask for a daily news brief with specific preferences and timing. Manage and customize your summary format and focus areas through the Tasks menu.
| Pro tip: Add specific topics you're interested in for more relevant news updates. |
|
| | PRESENTED BY FYXER | | | The Rundown: Fyxer AI is an executive assistant that handles email management and meeting notes autonomously, learning your communication style to deliver personalized support and help you reclaim an hour every day. | With Fyxer AI, you can: | Handle your inbox efficiently with AI-powered prioritization and response drafting — with 63% of drafted emails requiring zero edits Capture every meeting detail with automated summaries and action items Scale productivity across teams with adaptive AI assistance
| Try Fyxer AI free for 7 days with an easy 30-second setup with Gmail or Outlook. |
|
| | RUNWAY | |  | Image source: Runway |
| The Rundown: Runway just officially released its Frames AI image generation model, a new addition to the creative platform that offers high-quality photorealism and stylistic control through 'Worlds' that maintain consistent characteristics. | The details: | The model was initially revealed in November, alongside 10 sample 'Worlds' allowing users to maintain a specific aesthetic throughout generations. Outputs can be used in Runway's video tools or edited with changes like fixed seeds, a 'vary' action, and controls like aspect ratio, style, and aesthetic. Frames is available to paid users on both the Unlimited and Enterprise plans, with each generation costing 32 credits for four image outputs.
| Why it matters: While Runway was already a leader in video generation, having an image model that looks to be on par with the industry's best is quite the value add. With enhanced stylistic control, incredible realism, and easy plug-and-play into powerful video models, Runway unlocks powerful new creative workflows. |
|
| | | | | 🧊 Krea AI 3D - Turn any image into a 3D object to manipulate in other scenes 🗣️ MiniMax T2A-01 HD - Text-to-audio model enabling voice cloning with just 10 seconds of audio and support for over 17 languages. 📣 Humva - Create free customized avatars with thousands of templates for social media, product intros, and more 🔎 Get Website Report - AI agent that audits webpages for SEO and conversion
|
|
| | | | | Perplexity acquired professional networking platform Read.cv and is reportedly launching a merger bid for TikTok's U.S. operations valued at over $50B. | Epoch AI is facing backlash after OpenAI's financial backing of the Frontier Math benchmark was revealed belatedly following o3 achieving record-breaking performance. | New data from the Pew Research Center revealed that ChatGPT usage for schoolwork among U.S. teens has doubled to 26% since 2023, with over 79% of teens now having heard of the platform. | Character AI launched two new interactive games, making a strategic transition toward entertainment integrated with AI-powered chatbots. | Cognition Labs released version 1.2 of its Devin AI coding assistant, enhancing capabilities with improved context understanding, a browser-based workplace setup, new enterprise accounts, and Slack audio integration. |
|
| | | | | | Join our next workshop on Tuesday at 2:15 PM EST and learn how to build a complete, automated workflow for advanced meeting preparation using Lindy AI with Flo Crivello, Lindy's visionary founder and CEO. | RSVP here. Not a member yet? Join The Rundown University on a 14-day free trial. |
|
| | | We'll always keep this newsletter 100% free. To support our work, consider sharing The Rundown with your friends, and we'll send you more free goodies. | | Your current referral count: 0
What you need: 1 more referrals until receiving ChatGPT Prompting Kit Bundle | |
|
| | 👀 Get featured in the world's largest AI newsletter | Interested in sponsoring us and reaching 914,056 AI enthusiasts, tech executives, investors, engineers, and business owners around the world? Get in touch today. |
|
| | That's it for today!Before you go we'd love to know what you thought of today's newsletter to help us improve The Rundown experience for you. | | See you soon, | Rowan, Joey, Zach, Alvaro, and Jason—The Rundown's editorial team |
|
|
|
Tidak ada komentar:
Posting Komentar