⚡️ AI Safety Summit

Good Morning. The first AI Safety Summit has come to an end with Rishi Sunak hailing “landmark” agreements and progress on global collaboration around artificial intelligence. Let’s dive right in.

Today’s Highlights:

• Midjourney rolls out style tuner

• China’s new AI chip reportedly outperforms Nvidia

• Brave's new privacy-first AI chatbot

DEEP DIVE

First Global Gathering on AI Safety Concludes

The first global AI Safety Summit at Bletchley Park has wrapped, leading to expansive discussions on ethical AI and safety measures.

The summit, backed by several governments and major AI firms like OpenAI and Google DeepMind, facilitated agreements to test new AI models for possible harmful capacities, including national security and societal harms, before their release.

Elon Musk highlighted AI as "one of the biggest threats" to modern society since "we have for the first time the situation where we have something that is going to be far smarter than the smartest human," stressing the need for a "third-party referee" in the AI industry.

So, what does it all mean for the future of AI safety?

• New institutes for AI Safety will emerge in the UK and the US, tasked with the evaluation and testing of new AI models.

• These institutes, along with more to come, could form a network focusing on AI safety and possibly lay the groundwork for extensive AI regulations.

• The Bletchley Declaration, recognizing the potential risks of AI, was signed by all countries present, pledging efforts towards safe and accountable AI development.

However, these big steps aren't covering every concern just yet. The agreements made are currently voluntary, and critics question if political tensions might hinder international collaboration. Also, there's the concern about AI's impact on existing jobs, notably in publishing and administrative work, which is yet to be thoroughly addressed.

Keep reading here.

PUNCHLINES

Rolling with Runway: Runway's Gen-2 undergoes major improvements to both the fidelity and consistency of video generations.

Robbing Hood: New AI capable of insider trading and lying could take the 'Risk' out of Wall Street.

Visual Voyager: Midjourney rolls out style tuner, allowing users to apply their unique visual style to (possibly) all generations.

Turn Up the Pixels: Kaiber's new app helps artists generate music videos artists using generative AI.

Bot with Benefits: Brave's new AI chatbot, Leo, prioritizes user anonymity and privacy—challenging Bing and ChatGPT.

TLDR

China reportedly develops AI chip outperforming Nvidia: AI researchers at Beijing's Tsinghua University have developed an all-analog chip (ACCEL), performing 74.8 quadrillion operations per second per watt, making it more efficient than Nvidia's popular A100. The chip, combining electronic and photonic computing, has applications in wearable devices, autonomous vehicles, and industrial inspections.

Artificial Intelligence Improves Efficiency and Patient Care: Implementing AI can significantly reduce non-essential tasks in healthcare, relieving clinicians and boosting patient experience, says Kalle Conneryd-Lundgren, COO of Kry, a health-tech firm. With AI reducing administrative duties by about 70%, clinicians can focus more on patient care. This has led to a 20% increase in efficiency and more personalized care, resulting in happier patients and clinicians.

Cohere unveils Embed V3 for advanced language model applications: AI startup Cohere has launched Embed V3, aimed at enhancing semantic search and applications for LLMs. Perfect for retrieval augmented generation, Embed V3 is optimized for delivering improved document-query matches and can be configured for applications such as search, classification, and clustering. It even features a multilingual version, reducing the `1costs of running vector databases.

TRENDING TOOLS

🔍 Docus.ai: Rapid diagnosis with AI, backed by top medical experts

☁️ Empy.ai: Optimize team communication and prevent conflicts on Slack

🧞‍♂️ Genie by Luma Labs: Transform text into 3D models through research-level technology

🛠️ Freeplay: Revolutionize your construction process with LLMs

📊 Herbie by Broadn: Manage intricate business and marketing tasks

That’s all for today—if you have any questions or something interesting to share, please reply to this email. We’d love to hear from you!

P.S. If you want to sign up for the Supercharged newsletter or share it with a friend, you can find us here.