🤯AGI Test: Humans 60%, AI 1%

Today we will discuss:

• Sponsored: Innovating with AI—Helping AI enthusiasts become high-earning consultants with the tools, strategies, and clients to succeed 

• AI vs. humans: The intelligence gap—New AGI test reveals AI still struggles to think like us

• Don’t miss this week’s Workflow Wednesday—AI strategy unlocked with Claude 3, Google Gemini, and SEOwind;now available at only $20/month! 

• AI chip curbs put Trump on the spot—Industry leaders and diplomats urge policy shift

• In other AI news today — AI-generated audio falsely depicting JD Vance criticizing Elon Musk goes viral, OpenAI reshuffles leadership as Sam Altman shifts to a technical role, OpenAI’s AI voice assistant gets a more natural feel, and EU lawmakers warn against weakening AI regulations

• Saved the best for last—10 must-try AI tools

 💼Want to build a 6-figure business as an AI Consultant?

The AI consulting market is about to grow by a factor of 8X – from $6.9 billion today to $54.7 billion in 2032. But how does an AI enthusiast become an AI consultant? How well you answer that question makes the difference between just “having AI ideas” and being handsomely compensated for your contribution to an organization’s AI transformation.

Thankfully, you don’t have to go it alone – our friends at Innovating with AI have welcomed 600 new students into The AI Consultancy Project, their new program that trains you to build a business as an AI consultant. Some of the highlights current students are excited about:

• The tools and frameworks to find clients and deliver top-notch services

• A 6-month plan to build a 6-figure AI consulting business

• Students getting their first AI client in as little as 3 days

And as a reader of OpenTools.Ai, you have a chance to get early access to the next enrollment cycle.

Click here to request early access to The AI Consultancy Project

New AGI benchmark proves AI can’t think like humans yet

Key Points 

• ARC-AGI-2 challenges AI models with unseen pattern puzzles, testing their ability to adapt rather than memorize solutions.

• Leading AI models, including GPT-4.5 and Claude 3.7 Sonnet, scored around 1%, far below human participants’ 60% average.

🧠Context of the news - The Arc Prize Foundation, co-founded by AI researcher François Chollet, has introduced a new way to test artificial general intelligence. The ARC-AGI-2 test challenges AI models to recognize patterns in visual puzzles and generate solutions to problems they haven’t encountered before. 

🤖News - So far, the results suggest there’s still a long way to go. OpenAI’s o1-pro and DeepSeek’s R1, both designed for reasoning, scored between 1% and 1.3%. Other top models, including GPT-4.5, Claude 3.7 Sonnet, and Gemini 2.0 Flash, also landed around 1%.

🫡Humans outperform AI - In contrast, a group of over 400 human participants achieved an average score of 60%, far surpassing any AI model.

🤓Why this is more reliable - Chollet believes ARC-AGI-2 is a more accurate measure of intelligence than its predecessor, ARC-AGI-1, because it prevents AI from using brute-force computing power to find answers. The test introduces a new efficiency metric and requires models to interpret patterns in real time instead of relying on memorization.

👨‍💻Why it matters - As AI advances, researchers are searching for better ways to measure its capabilities. Chollet and others believe that AI’s ability to efficiently acquire new skills, not just solve tasks, will be a key factor in determining true artificial general intelligence.

Read the full story here. 

We just launched a new edition of Workflow Wednesday for AI-minded professionals like you. Actionable AI-workflows delivered directly to your inbox.  This week's Topic? AI-Monetization

Join Here

Trump faces growing calls to rework AI chip export rules

Key Points 

• The AI diffusion rule limits chip exports, sparking opposition from Nvidia, Oracle, and US allies like Israel and Poland.

• With a compliance deadline approaching, Trump officials remain divided on whether to ease restrictions or keep them in place.

• If unchanged, the rule could reshape global AI infrastructure and turn US chip access into a geopolitical bargaining tool.

☕News - The Trump administration is under growing pressure from foreign officials and major tech companies to reconsider its approach to AI chip exports. 

At the center of the debate is the AI diffusion rule, a policy restricting how many AI processors can be sent to most countries. First introduced in the final days of the Biden administration, the rule has drawn criticism from companies like Nvidia and concerns from US allies such as Israel and Poland, who worry it could disrupt chip supplies and deter AI investments.

🤹‍♂️Possible changes - With a compliance deadline just two months away, governments and companies are urging Trump officials to ease the restrictions. Inside the administration, however, there’s no clear consensus. While a full repeal isn’t being considered, some officials have suggested revising the rule by scrapping its tiered structure or adjusting how export licenses are handled.

🥸What's at stake - If the policy stays in place as written, it could reshape AI infrastructure plans in key global markets and turn access to US technology into a geopolitical bargaining tool. 

For companies like Oracle and Nvidia, which are expanding overseas, maintaining any version of these restrictions would be an unwelcome hurdle. 

Meanwhile, Washington’s broader goal remains unchanged: keeping AI development tied to the US and its closest partners while ensuring AI infrastructure abroad meets American security standards.

Read the full story here.