Welcome to AI Box, your weekly source for the latest developments in AI technology and applications.

Today's AI News lineup:🥊

• 🤖 Geekbench AI 1.0 Launches, Setting New Standards

• 🎨 Grok Unleashes Unfiltered AI Image-Generation with Black Forest Labs

• 🧠 New Study Highlights Persistent AI Hallucinations

🤖 Geekbench AI 1.0 Launches, Setting New Standards

AI Box: Primate Labs has released Geekbench AI 1.0, a comprehensive benchmark tool for assessing AI performance across multiple platforms.

Key Highlights:

• Geekbench AI 1.0 expands on Geekbench ML, providing a standardized performance measure for machine learning and AI workloads.

• The update reflects a broader industry shift towards standardized AI benchmarking and clearer performance metrics.

• OpenAI also introduced its own benchmark, SWE-bench Verified, which uses human validation to assess models' real-world problem-solving effectiveness.

Why It Matters: As AI technology continues to evolve, standardized benchmarks like Geekbench AI 1.0 and human-validated tools like SWE-bench Verified are essential for ensuring accurate performance comparisons and practical efficacy in real-world applications.

🎨 Grok Unleashes Unfiltered AI Image-Generation with Black Forest Labs

AI Box: Elon Musk’s Grok, in collaboration with Black Forest Labs, has introduced a new AI image-generation feature with minimal safeguards, resulting in a flood of controversial images on X.

Key Highlights:

• Grok's new AI image-generator, powered by Black Forest Labs' FLUX.1 model, allows users to create and share highly controversial and unregulated images on X.

• Black Forest Labs, a recently launched startup, aims to push the boundaries of AI image generation, surpassing established models like Midjourney and Dall-E.

• The lack of safeguards in Grok's image generator has led to a surge in misleading and potentially harmful images, raising concerns about misinformation.

Why It Matters: The introduction of Grok’s unregulated AI image generator exemplifies the tension between innovation and ethical considerations in AI, highlighting the risks of spreading misinformation and the broader implications for digital media platforms.

🧠 New Study Highlights Persistent AI Hallucinations

AI Box: A recent study reveals that despite advancements, generative AI models like GPT-4o and others still frequently produce hallucinations, with no model achieving consistent factual accuracy.

Key Highlights:

• Researchers from Cornell and other institutions found that generative AI models, including GPT-4o, often generate hallucinations, with factual accuracy only about 35% of the time.

• Models struggle most with questions outside Wikipedia and on complex topics like finance and pop culture, while geography and computer science questions are easier.

• Larger models and those with web search capabilities didn’t significantly outperform smaller models, suggesting that model size alone doesn’t mitigate hallucinations.

Why It Matters: This study underscores the ongoing challenge of ensuring reliable AI outputs, highlighting the need for improved fact-checking mechanisms and cautious use of AI in critical contexts.

💻 AI Box: No Code AI App Builder & Marketplace

AI Box is a no-code, AI app building platform paired with the App Store for AI that lets you monetize your AI tools.

The platform lets you build apps by linking together hundreds of AI models like ChatGPT, MidJourney, and Eleven Labs. Eventually we’ll integrate software like Gmail, Trello and Salesforce so you can use AI to automate every function within your organization.

To get notified when we launch and be the first to build on the platform just stay subscribed to this newsletter alternatively you can join the waitlist at AIBox.ai

Interested in sponsoring The AI Box Newsletter or the AI Chat Podcast?
Email [email protected] to see if you'd be a good fit!

#️⃣ AI Twitter

Sophia back with another joke!

That's all for today!

If you have any interesting projects or ideas, please reach out to us by responding to this email or by sending us a DM on Twitter: @jaeden_ai & @aibox_ai

As always, thanks for reading, and see you next time. 🫡