China’s latest AI tops GPT-4o

Welcome, AI enthusiasts.

Alibaba just unveiled Qwen2-VL, a brand new vision-language AI that outperformed GPT-4o across several key benchmarks.

As China’s AI developments proceed to speed up, is the gap between U.S. leaders closing faster than expected? Let’s get into it…

In today’s AI rundown:

• China’s latest Qwen2 beats GPT-4o

• OpenAI and Anthropic partner with US gov

• Add yourself to photographs with a custom LoRA

• AI startup reaches 100M token context

• 5 latest AI tools & 4 latest AI jobs

• More AI & tech news

Read time: 4 minutes

LATEST DEVELOPMENTS

ALIBABA

🤖 China’s latest Qwen2 beats GPT-4o

The Rundown: Alibaba just unveiled Qwen2-VL, a brand new vision-language AI model that outperforms GPT-4o in several benchmarks — particularly excelling in document comprehension and multilingual text-image understanding.

The small print:

• Qwen2-VL can understand images of assorted resolutions and ratios, in addition to videos over 20 minutes long.

• The model excels particularly at complex tasks equivalent to college-level problem-solving, mathematical reasoning, and document evaluation.

• It also supports multilingual text understanding in images, including most European languages, Japanese, Korean, Arabic, and Vietnamese.

• You may try Qwen2-VL on Hugging Face, with more information on the official announcement blog.

Why it matters: There’s one more latest contender within the state-of-the-art AI model arena, and it comes from China’s Alibaba. Qwen2-VL’s ability to know diverse visual inputs and multilingual requests could lead on to more sophisticated, globally accessible AI applications.

TOGETHER WITH ANYSCALE

🤝 The world’s largest gathering of AI leaders

The Rundown: Ray Summit 2024 is just across the corner. Get insights from cutting-edge AI corporations like NVIDIA, Intel, Google, and others. Featuring 60+ breakout sessions and hands-on tech training, only available in person in San Francisco from September 30-October 2.

At Ray Summit you’ll:

• Learn from OpenAI’s CTO, Instacart’s co-founder, and others

• Explore real-world AI solutions from tech giants using Ray and Anyscale

• Boost your skills with hands-on training in GenAI, LLMs, and more

Register now with code “Newsletter15” for an exclusive discount for readers of The Rundown.

OPENAI & ANTHROPIC

🏛️ OpenAI and Anthropic partner with US gov

The Rundown: OpenAI and Anthropic just signed a groundbreaking agreement with the U.S. Artificial Intelligence Safety Institute to permit government access and testing of their AI models before public release.

The small print:

• The U.S. AI Safety Institute can have access to major latest models from each corporations prior to and after their public release.

• This collaboration is a step toward AI regulation and safety efforts, with the U.S. government evaluating AI models’ capabilities and associated risks.

• The institute will provide feedback to OpenAI and Anthropic on potential safety improvements that needs to be made.

• These agreements come as AI corporations face increasing regulatory scrutiny, with California legislators recently passing a broad AI regulation bill earlier today.

Why it matters: The 2 hottest AI corporations on the earth are granting the U.S. government access to unreleased models before release. This might reshape how AI is developed, tested, and deployed worldwide, with major implications around innovation, safety, and international competition within the AI space, for higher or worse.

AI TRAINING

🖼️ Add yourself to photographs with a custom LoRA

The Rundown: Fal AI’s Flux LoRA training tool helps you create a customized AI image generation model that may create images of you in any scenario or style from a couple of selfies.

Step-by-step:

1. Visit Fal AI’s Flux LoRA training page and create an account (requires ~$10 in credits).

2. Upload 6-12 high-quality images of yourself with clear backgrounds.

3. Set training steps to 1000 and add a novel trigger word (e.g., “YourName”).

4. Start the training process (takes about 20 minutes).

5. Generate images using prompts like “Portrait of [YourName] as a superhero” and explore!

Pro tip: To go one step further, import your images into Runway’s Gen-3 image-to-video feature and switch your generations into short clips.

THE RUNDOWN AI UNIVERSITY

 📧 Create AI Agents that reply to your emails

The Rundown: Tomorrow we’re hosting a live workshop with Flo Crivello, the founding father of Lindy AI, to indicate you the right way to create AI agents that reply to emails, negotiate, and schedule meetings in your behalf.

Join us on live at 1:30 PM PST to learn:

• What Lindy’s AI agents are able to, and the right way to use the platform

• How one can create your personal agent that responds to emails in your behalf

• Exclusive Q&A for all members to ask any questions

For those who’re a member of The Rundown AI University, you possibly can RSVP here.

For those who’re not a member yet, you possibly can still join the workshop without spending a dime with a 14-day free trial.

MAGIC

📈 AI startup reaches 100M token context

The Rundown: Magic just developed LTM-2-mini, a model able to processing 100 million tokens of context — reminiscent of about 10 million lines of code or 750 novels — and partnered with Google Cloud to construct advanced AI supercomputers.

The small print:

• LTM-2-mini can process and understand 100 million tokens of context given during inference, surpassing current models by 50x.

• The model’s revolutionary algorithm processes long sequences of information 1000x more efficiently than the present top-performing AI models.

• Magic can be partnering with Google Cloud to construct supercomputers powered by Nvidia’s newest and most advanced GPUs.

• The corporate has raised greater than $450 million in total funding, including a recent $320 million investment round.

Why it matters: This breakthrough in context length allows AI agents to process and reason over dense and complex codebases, vast databases, and years of conversation history in a single inference. It’s a big step toward creating AI assistants with near-perfect recall and memory.

NEW TOOLS & JOBS

Trending AI Tools

• 💻 GPT Engineer – Chat with AI to construct web applications in your behalf

• 🚀 Mimrr – Automate documentation and evaluation for development teams

• 🤖 AgentOps – Develop and debug AI agents efficiently

• 🎨 Krea FLUX Style Mixer – Mix multiple Flux image generation styles with full control

• 🔎 Next Alpha Andromeda – LLM for crypto research

Latest AI Job Opportunities

• 🧑‍🔬 Mistral AI – AI Scientist – Internship

• 🤝 Meta – Business Development and Partnership Manager Artificial Intelligence

• 🚦Waymo – Product Manager, Perception

• 💪 Captions – Technical Recruiter

QUICK HITS

Meta reported significant growth for its Llama AI models, with downloads approaching 350 million and usage increasing 10x since January.

Nous Research released the Hermes Function Calling V1 dataset for training AI models in function calling and structured output capabilities.

Nvidia and Apple reportedly discussed joining OpenAI’s funding round with Microsoft, potentially valuing the AI startup at over $100 billion.

California lawmakers approved a bill proposing sweeping AI regulations, including safety testing requirements and potential legal consequences for harmful AI systems.

Yale University announced a $150 million investment over 5 years to support AI research, development, and education across the institution.

Codeium raised $150 million in Series C funding, reaching a $1.25 billion valuation and unicorn status lower than two years after launch.

Playground launched a brand new AI-powered graphic design tool allowing users to make logos, social media and t-shirt designs, and more without spending a dime.

THAT’S A WRAP

SPONSOR US

Get your product in front of over 650k+ AI enthusiasts

Our newsletter is read by hundreds of tech executives, investors, engineers, managers, and business owners all over the world. Get in contact today.

FEEDBACK

If you will have specific feedback or anything interesting you’d wish to share, please tell us by replying to this email.