Newsletter #94 - AI-generated content

AI-generated content is here to stay

Three years ago, a digital creative director of a Japanese ad agency wanted to find out who could better direct a TV commercial: humans or AI. The result? Ad execs preferred the robot's creative direction—were you surprised?

Welcome to the age of AI-generated content, where robots can write films and can seemingly make arguments, as seen in this AI-written opinion piece published in The Guardian.

About 37% of businesses are using AI in some shape or form, according to a Gartner study. If it's possible, should businesses and us let AI do all of the hard work? What's left for us humans?

Outlearning the robots,
Team NC

🤖 AI for content creation: friend or foe?

#TRENDS

Robo-writers: rise and risks of language-generating AI [Nature]

A new AI that dazzled Silicon Valley - Called GPT-3, the AI generate fluent streams of text that after imbibing billions of words from books, articles and websites was created by OpenAI in San Francisco a year ago. GPT-3 was so good that people found it hard to distinguish its news stories from prose written by humans.

Use cases for business - Large language models are already business propositions. Google uses them to improve its search results and language translation; Facebook, Microsoft and Nvidia are among other tech firms that make them. OpenAI keeps GPT-3’s code secret and offers access to it as a commercial service.

Dangers and solutions - Despite its versatility and scale, GPT-3 still has serious weaknesses like spewing hate speech and generating racist and sexist stereotypes.

One way to solve the bias is to weed out toxic text from the training data. But even if we agreed on what counts as toxic, Amanda Askell, philosopher and research scientist at OpenAI, says we might not want to blind language models. She adds: “If I had a model that had never had any exposure to sexism, and you were to ask it, ‘Is there any sexism in the world,’ maybe it just says, ‘no’.”

#BUSINESS

Microsoft is giving businesses access to OpenAI’s powerful AI [The Verge]

Repackaging AI for businesses - Microsoft is partnering with OpenAI to make an upgraded version of the GPT-3 AI language model and is making it available to Microsoft's business customers, where the AI can be used as a autocomplete tool for text-based content such as emails.

What are the applications? - Apart from autocompleting snippets of text, the AI's ability to parse language allows it to take on other tasks like summarising documents and generating ideas for projects and stories.

Here’s an example scenario from Microsoft: “A sports franchise could build an app for fans that offers commentary and a summary of game highlights, lowlights and analysis in real-time. Their marketing team could then use GPT-3’s capability to produce original content and help them brainstorm ideas for social media or blog posts and engage with fans more quickly.”

Word of caution - Some experts note that Microsoft’s introduction of GPT-3 fails to meet the company’s own AI ethics guidelines. Nonetheless, Microsoft's AI can avoid a lot of trouble by simply selecting its customers carefully and thus selecting what the AI will be specifically used for.

#TECH

How AI will help future artists make better visual content [Venture Beat]

AI for the creator economy - The vast majority of content online is visual content. According to a new report by LDV Capital, there will be at least 100x more visual content in the world by 2027, and the future creator economy will be powered by AI tools that will automate various aspects of content creation.

Predictions & possibilities - Whether it’s the army of the dead in Game of Thrones or a resized image of Kim Kardashian in a magazine, they're still content created by human artists. Now, computers and artificial intelligence can generate images and videos of people, things and places that never physically existed. By 2027, we will see more photorealistic synthetic images and videos than ones that document a real person or place.

Machine learning will also help curate, authenticate, and moderate content. As creators compete to produce greater volumes of digital content, a variety of tools (like this) will be developed to automatically repackage content from long-form to short-form, from videos to blogs, or vice versa, social posts, and more.

🤙 More handpicked reads (by a human)

Other Newsletters

Newsletter #77 - Billionaire space race

What the billionaire space race means for rest of us

Newsletter #74 - AI and big data

How AI and big data are fuelling APAC’s retail industry

Newsletter #71 - Electric cars

China tech titans bet $19bn on electric car frenzy

Newsletter #107 - How SEA is gearing up for the EV revolution

Evergrande gets go-ahead to sell electric vehicles, Tesla's self-driving beta disaster; how to find purpose

Newsletter #108 - Future of language AI

What if we could also build machines that can understand and use language just like we do?
View All Newsletters >