A $20 million AI competition

Hello, friends! Today is a good day for AI news, and we have the latest and greatest updates for you. 📰

🔨 AI TOOLS OF THE DAY

🎥 Papercup: Automate video dubbing with AI in multiple languages.→ check it out

😎 AI Social Bio: Automatically generate your social media bio. → check it out

📸 RestorePhotos: Restore old and blurry photos with AI. → check it out

ANTHROPIC’S NEW COLLABORATION 🤝

Like French fries and ketchup, tech companies are coming together to transform the taste buds of AI systems. 🍟

What’s up? AI developer Anthropic is poised to receive a substantial investment boost of $100 million from the prominent South Korean telecommunications powerhouse, SK Telecom.

Why? The primary objective of this initiative is to empower the creation of tailor-made AI services for global telecom service providers, underscoring the drive towards industry-specific AI solutions. Dario Amodei, the CEO of Anthropic, said, “We see industry-specific LLMs as having high potential to create safer and more reliable deployments of AI technology.”

More than just money? The collaboration between Anthropic and SK Telecom goes beyond mere financial backing, as the companies have additionally decided to collaborate on a multilingual large language model (LLM) for the Telco AI Platform. This upcoming LLM is slated to support languages including Korean, English, German, Japanese, Arabic, and Spanish. Anthropic and SK Telecom are also collaborating on the refinement of Anthropic's latest model, Claude 2.

THE NEW YORK TIMES PROHIBITS AI ACCESS 🚫

The New York Times is tired of data scraping … 🤛

What happened? The New York Times has revised its Terms of Service to explicitly prevent its content, spanning text, visuals, multimedia, metadata, and more from being employed in the development of any software, including machine learning and AI systems.

More details? The updated terms also stipulate that automated tools such as web crawlers, intended for accessing or aggregating this content, cannot be used without prior written consent from The New York Times. Non-compliance with these new guidelines may result in undisclosed fines or penalties.

Why? This move might be a response to Google's recent modification of its privacy policy, granting itself permission to train its AI services using publicly accessible web data. Given that many influential AI models, including OpenAI's ChatGPT, rely on extensive datasets that might incorporate copyrighted materials scraped from the web, this action seems aimed at addressing potential copyright concerns and protecting human authors.

THE U.S. GOVERNMENT INTRODUCES AN AI COMPETITION 🦾

A new competition has been introduced, and it’s a battle to the death! … Well, not exactly. ⚔️

What’s up? A competition has been initiated by the United States to create AI systems dedicated to identifying and resolving security concerns within the nation’s government infrastructure.

Why? Recent years have witnessed various U.S. sectors falling prey to hacking, underscoring the severity of cyber threats, especially from foreign actors. Canada's cybersecurity chief, Sami Khoury, echoed this concern, observing that AI is being exploited to craft phishing emails, propagate disinformation, and compose malicious code. As Anne Neuberger, the U.S. government's deputy national security advisor for cyber and emerging technology, said, "Cybersecurity is a race between offense and defense." She hopes the competition can help bring this idea to life.

How will this work? Major technology players such as Google, Anthropic, Microsoft, and OpenAI are contributing their AI systems to the competition, emphasizing collaborative efforts to bolster cybersecurity. Spearheaded by the Defense Advanced Research Projects Agency (DARPA), the two-year competition carries a prize pool of approximately $20 million. 🤑

TWITTER TUESDAY 🐦

HAS AI REACHED SINGULARITY? CHECK OUT THE FRY METER BELOW