• Data Talent Pulse
  • Posts
  • 👩‍💻 EU passes AI act, New nonprofit spotlights “Fairly Trained” AI

👩‍💻 EU passes AI act, New nonprofit spotlights “Fairly Trained” AI

DTP #43: The Race for AI Talent Heats Up

💼 AI in Business 

What Every CEO Needs To Know About The New AI Act 

According to an article from Forbes, The European Union has passed the Artificial Intelligence Act, implementing strict regulations on AI. 

  • Unacceptable AI applications are banned, except for specific government, law enforcement, and scientific purposes. 

  • Similar to GDPR, this law affects any business operating in the EU, regardless of location. 

  • The aim is to protect citizens' rights while promoting innovation. 

  • Enforcement could start in six months, with penalties for breaches up to €30 million or 6% of global turnover. 

  • Banned uses include influencing harmful behaviors, biometric classification for sensitive traits, and discriminatory social scoring. 

  • High-risk AI, like self-driving cars and medical applications, face stricter rules. 

  • Transparency requirements apply, including marking AI-generated content and disclosing information on AI models. 

  • Future AI regulation is expected globally, requiring businesses to understand risk and maintain ethical AI practices. 

  • Compliance and transparency are crucial for businesses to adapt to evolving AI regulations. 

👩‍💻 The Race for AI Talent Heats Up 

The Financial Times takes a look at the moves big tech is making in the battle for AI talent:  

  • Microsoft's latest move involves hiring talent from Inflection, leaving the company behind, showcasing a unique approach in the talent war. 

  • Mustafa Suleyman and Karén Simonyan, founders of Inflection, are now overseeing Microsoft AI, indicating Microsoft's intent to dominate the AI market. 

  • There's an intense talent war for AI experts, with companies like Microsoft, Google, and Nvidia offering lucrative salaries and benefits. 

  • Despite the dominance of big tech, some AI experts prefer working for startups or traditional companies outside the US. 

  • There's a global trend of AI talent moving to countries like the UAE, Saudi Arabia, and South Korea, as well as to traditional industries like healthcare and manufacturing. 

  • The AI world is deepening and diversifying, with smaller companies collectively hiring more top AI talent than the big five combined. 

🌐 From the Web 

AI is evolving search, with tools like ChatGPT, Google Gemini, and Microsoft Copilot. Companies like Perplexity and You.com aim to revolutionize search, but face challenges in replacing Google's multifaceted functionality and speed.  

Databricks unveiled DBRX, surpassing Meta's Llama 2 as the most potent open-source AI model. Costing $10 million to develop, it excels in various benchmarks, signaling a significant advancement in AI. 

As AI advances, economists predict real inflation-adjusted interest rates will rise due to increased demand for capital expenditures and higher productivity, diverging from conventional wisdom. 

☮ AI for Good 

New Nonprofit Spotlights AI Trained on Copyrighted Work with Permission 

Ed Newton-Rex, a computer scientist, departed from Stability AI due to concerns about the company's use of copyrighted material without artists' permission. This led him to establish Fairly Trained, a nonprofit organization aimed at certifying companies that ethically train their generative AI models using licensed data.  

  • Fairly Trained's certification process ensures companies use authorized training data and adhere to ethical standards. 

  • Newton-Rex intentionally focuses on smaller AI companies for their agility in adopting ethical practices. 

  • The certification process involves a written submission detailing training data sources and internal processes for compliance. 

  • Newton-Rex acknowledges the process relies on trust initially but effectively distinguishes ethical companies. 

  • Certified models primarily involve music, audio, or images due to challenges in text generation. 

  • Newton-Rex reflects on past use of generative AI for music creation, expressing hesitation considering the evolving ethical landscape. 

  • He emphasizes the potential for generative AI as assistive technology, democratizing creativity while respecting data creators. 

  • Newton-Rex advocates for a future where generative AI harmoniously coexists with human artists, emphasizing the importance of acknowledging and compensating creators. 

🤖 Prompt of the week 

Act as a deep learning expert. I have a dataset [describe dataset]. Write code to perform transfer learning using a pretrained model from TensorFlow Hub. 

See you next week, 

Mukundan 

The Data Talent Pulse is brought to you by TeamEpic, a trusted global AI Talent provider. Learn more about us here. 

Join the conversation

or to participate.