Welcome, humans.☕

What to decode:

• 👩‍🎤 Famous musicians are divided over whether to embrace or reject AI

• 📱 Runway release their AI video generation app

• 🧑‍💻 AI-powered customer support: The future of customer service

Read time: 4 minutes

👩‍🎤 Famous musicians are split on AI

The use of AI has sparked a debate among famous musicians, who are divided over whether to embrace or reject its involvement in the music industry. Recently, AI was used to create a fake Drake song that went viral, causing controversy and legal issues for Universal Music Group.

Grimes, a well-known artist, has even offered to split royalties with anyone who can create an AI version of herself. While some musicians are excited about the possibilities AI could bring to music, others have concerns that need to be addressed. Semafor takes a closer look at the legal issues surrounding AI in music. Read here

 📱Runway release their AI video generation app

If you're interested in exploring the video aspect of AI art creation, Runway appears to be a promising contender to follow. They have been consistently associated with all things related to video. Runway has recently launched a commercial app named RunwayML, which provides a platform for users to interact with AI-generated videos.

Currently, the app allows users to create some unique and eerie edits of their videos, but soon it will also support text-to-video generation. However, there is a caveat: the app offers limited free credits, and it can be quite expensive beyond that.

 🧑‍💻 AI-powered customer support: The future of customer service

AI in customer support is a point of agreement in a world filled with disputes. According to researchers from Stanford/MIT, AI has the potential to revolutionize customer support and reduce the time spent on hold.

A software company introduced AI to over 5,000 customer support agents, resulting in a 14% increase in overall productivity. The most significant productivity gains were observed among junior workers, who completed their tasks 35% faster than usual. However, top-performing agents who had already developed their own tools saw little improvement.

Agent productivity also increased more quickly with the use of AI. Agents who had been using AI for only two months performed just as well as those with six months of experience and no AI. The key takeaway is that the AI model learned the best practices from the top performers and made them available to the entire group, thus elevating every worker to the status of 'Employee of the Month.'

There is currently a global race to deploy AI in support systems. Perhaps Meta, which owes a response to a six-week-old support ticket, will finally provide an answer.

FASCINATING FINDS

• On this day - Bell Labs demonstrated the first practical silicon solar cell in New Jersey, US. (link)

• Weird but interesting - A water-powered computer. In 1936, Vladimir Lukyanov built a computer that used water to solve partial differential equations. The computer was called a "Water Integrator" and was originally designed to solve the problem of cracking in concrete. It is now on display in the Polytechnic Museum in Moscow. (link)

• Mind-boggling statistics - The amount of data created every day is now estimated to be 2.5 quintillion bytes, or 2.5 billion gigabytes. This is equivalent to the amount of data that would be stored on 100 million DVDs.(link)

MISCELLANEOUS

(News, podcasts, videos, blogs etc)

• Greywing’s new SeaGPT solves email overwhelm for maritime crew managers. (link)

• Solving CTF hacking challenges with the help of GTP-4. (link)

• The future of AI relies on a high school teacher’s free database. (link, without the paywall here)

• Russia's Sberbank launches own version of ChatGPT. (link)

• The unbearable heaviness of being positioned. (link)

• Google brings generative AI to cybersecurity with Sec-PALM. (link)

• We need to become less dependent on OpenAI’s models. (link)

• GitLab’s new security feature uses AI to explain vulnerabilities to developers. (link)

• Why I'm not worried about AI causing mass unemployment. (link)

That's all for now!

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