AI Agents That Build New AI Agents in Real-Time 🤯

Welcome to another edition of Horizon AI,

A new startup has emerged from stealth with a fresh take on AI agents, using them to create more agents that can perform more complicated tasks on behalf of users.

Let’s get started!

Read Time: 4.5 min

Here's what's new today in the Horizon AI

• Emergence AI Is Using AI Agents to Build New AI Agents in Real-Time

• OpenAI's ChatGPT Subscription Surpasses 20 Million Users

• AI Tutorial: How to analyze YouTube videos with Gemini

• AI Tools to check out

• The Latest in AI and Tech 💡

• AI Findings/Resources

AI News

EMERGENCE AI

Emergence AI Is Using AI Agents to Build New AI Agents in Real-Time

Emergence AI has announced a new AI agent creation platform that lets the user specify what work they are trying to accomplish via text prompts, and then turns it over to AI models to create the agents they believe are necessary to accomplish said work.

Details:

• The new system is a no-code, natural language, AI-powered multi-agent builder that works in real time.

• When given a task, Emergence AI evaluates the required work and checks its registry of AI agents for the right fit. If none are suitable, it autonomously creates new agents to handle the task.

• The platform can also proactively create agents to anticipate related tasks based on the user’s previous requests and coordinate multiple agents to work together in a fully autonomous way, enabling a new level of enterprise automation.

• The startup claim its platform represents a new milestone in the area of “recursive intelligence,” which refers to AI systems that can improve themselves over time without human intervention.

• The CEO, Satya Nitta, explains that recursive intelligence is the key factor that enables AI agents to create new AI agents. 'Our systems allow creativity and intelligence to scale fluidly without human bottlenecks, but always within human-defined boundaries.'

Emergence AI describes its work as being in the early stages and has not disclosed pricing for its platform. However, it invites anyone interested in the concept to contact the company directly for details on access and costs.

Source

TOGETHER WITH SQUARESPACE

Design Intelligence: Squarespace’s Approach to AI

Through Squarespace’s cutting-edge features that combine automation, design presets, creative guidance, and generative AI, Design Intelligence makes it easy to build a beautiful and impactful website. With just a few pieces of information, Blueprint AI generates an entire website customized based off your brand’s goals, name, and personality.

Each page is populated with beautiful, AI-generated imagery that’s been carefully designed with a proprietary prompting approach and custom art direction, as well as generative text copy that matches your website’s category and tone. Each image is beautiful and inspirational, and each bit of text provides a helpful jumping off point. It’s AI speed, with Squarespace’s 20+ years of design expertise in website building.

Use code SAVE10 for 10% off today!

OPENAI

OpenAI's ChatGPT Subscription Surpasses 20 Million Users

OpenAI saw its revenue climb 30% in just three months as ChatGPT's paid subscribers surpassed 20 million, up from 15.5 million at the end of 2024.

Details:

• With the subscriber increase, OpenAI is now generating at least $415 million in monthly revenue, up from $333 million late last year, putting it on track for $5 billion in annual revenue.

• The platform's total weekly user base grew even faster, reaching 500 million users, likely fueled by the recent hype around Ghibli-style image generation. The proportion of paying users fell from 5% to 4%.

• The figures follow news of OpenAI's $40 billion funding round, the most money raised in a single round by a private tech company.

Despite strong revenue growth, profitability remains elusive as the company increases investments and serves more free users, leading OpenAI CEO Sam Altman to express concerns about capacity issues and warn that these constraints could cause delayed product launches and service disruptions.

Source

AI Tutorial

How to analyze YouTube videos with Gemini

1. Go to Gemini’s website and sign in.

2. Choose ‘Gemini 2.0 Flash Thinking Experimental’ as the model

3. Go to Settings > Apps > Media, and enable ‘YouTube’.

4. Copy the URL of the YouTube video, paste it into Gemini, and enter your prompt.

Sample Prompt: ‘Analyze this video and give me key insights and practical applications of the concepts explained. Include timestamps for easy reference [enter video URL]’

You'll receive a quick, detailed analysis of the video in just a few seconds.

AI Tools to check out

⭐ Wepost: Simplify your social media workflow.

🎮 Focumon: Gamify your daily goals.

🔎 Picurious: Get instant answers, insights, and inquisitive questions about your pictures, and easily identify any object, plant, or text.

✅ Dify: An open-source LLM app development platform.

✨ Hotpot: Create amazing images, graphics, and writing with AI.

AI Findings/Resources

👀 Researchers warn that more pre-training data may not always lead to better models

🤔 An AI bubble threatens Silicon Valley, and all of us

👉 10 professional developers on vibe coding's true promise and peril

🔮 Technology has shaped human knowledge for centuries. Generative AI is set to transform it yet again

The latest in AI and Tech

Tinder's new game tests your flirting skills with AI personas powered by OpenAI

The Game Game is an experience designed to help you practice your conversation skills by talking to an AI voice, powered by OpenAI's GPT-4o model.

It works by dealing you a "stack of cards," each card containing a different AI persona and "meet cute" scenario, and the goal is to score a date or a phone number within the time limit.

Meta's head of AI research stepping down

Meta’s VP of AI research, Joelle Pineau, announced in a post on Facebook that she is planning to leave the company in May after more than two years overseeing FAIR, Meta’s internal AI research lab led by Yann LeCun.

OpenAI launches free AI courses for learners worldwide

OpenAI has quietly launched the OpenAI Academy, a free educational platform offering dozens of hours of video content and live events designed to teach AI fundamentals and practical skills, with no payment or prerequisites required to join.

Meta develops AI system MoCha that turns text into animated characters

Meta and University of Waterloo researchers have built MoCha, an AI system that generates complete character animations with synchronized speech and natural movements.

MoCha outperformed other systems in lip-sync quality and movement in tests, but the company hasn't revealed whether the system will become open source or remain a research prototype.

That’s a wrap!

Thanks for sticking with us to the end! Let’s stay connected on LinkedIn and Twitter.

Not subscribed yet? Sign up here and send it to a colleague or friend!

See you in our next edition!

Gina 👩🏻‍💻