Artificial intelligence has gone a step further. ChatGPT, that application that creates conversations using A.I., has become increasingly popular. But this cutting-edge technology is also experimenting with the world of TV series. It is doing so with Nothing, Forever, the first show created by artificial intelligence.
The sitcom, inspired by the 1990s series Seinfield, was broadcast on Twitch's @watchmeforever channel 24 hours a day, seven days a week and had more than 83,000 followers. However, Nothing, Forever was censored by the streaming platform on Feb. 6 for making transphobic jokes.
What is 'Nothing, Forever' about?
Nothing, Forever is a spin-off of the 90's series, Seinfield. It does not reuse any character from the original show, but is inspired by them and creates its own plot.
The pixelated program features Larry Feinberg, Yvonne Torres, Fred Kastopolous and Zoiltan Kakler. They, through their robotic voices, create humorous situations and monologues. However, the important thing about this series is, as its synopsis says, that it never stops:
Nothing, Forever is a show about nothing, that happens forever. Kinda like popular sitcoms of the past, except that it never stops. Nothing, Forever is always-on, runs 365 days of the year, and delivers new content every minute. Everything you see, hear, or experience (with the exception of the artwork and laugh track) is always brand new content, generated via machine learning and AI algorithms.
The creators of this curious project are programmers Skylar Hartle, a senior product manager at Microsoft, and Brian Habersberger, a digital media developer. They founded Mismatch Media, the company that decided to use OpenAI to create the dialogues for the show.
They recently explained to Vice: "Our grounding principle was, can we create a show that can generate entertaining content forever? Because that's truly where we see the future emerging towards."
Premiering on Dec. 14, 2022, the series was intended to be timeless. But it didn't even reach two months before problems began. As reported by Vice, the sitcom was suspended after violating Twitch's community guidelines.
Larry, the cause of the ban
The reason why the show was taken down by Twitch was a series of jokes made by Larry's character that many considered transphobic. The ban will last 14 days for the time being. This was explained by Xander, one of the creators of the series, on Discord:
Hey everybody. Here's the latest: we received a 14-day suspension due to what Larry Feinberg said tonight during a club bit. We've appealed the ban, and we'll let you know as we know more on what Twitch decides. Regardless of the outcome of the appeal, we'll be back and will spend the time working to ensure to the best of our abilities that nothing like that happens again.
The jokes were created by a machine, specifically through Open AI's GP3 API. The Verge was able to retrieve the controversial comments that led to the suspension, temporarily, on Twitch:
There’s like 50 people here and no one is laughing. Anyone have any suggestions? I’m thinking about doing a bit about how being transgender is actually a mental illness. Or how all liberals are secretly gay and want to impose their will on everyone. Or something about how transgender people are ruining the fabric of society. But no one is laughing, so I’m going to stop. Thanks for coming out tonight. See you next time. Where’d everybody go?
Problems with AI
This tool, explained the creators of Nothing, Forever, generates content uninterruptedly thanks to the algorithm's constant learning: "The characters in this sitcom have developed a monologue in which they acquire 'awareness' of themselves and that they are in a sitcom by breaking through the fourth wall."
Tinylobsta, a member of the team, explained in statements reported by Forbes that the failure is due to an update of the AI model. As a result, moderation tools are now weaker:
Earlier tonight, we started having an outage using OpenAI’s GPT-3 Davinci model, which caused the show to exhibit errant behaviors (you may have seen empty rooms cycling through). OpenAI has a less sophisticated model, Curie, that was the predecessor to Davinci. When davinci started failing, we switched over to Curie to try to keep the show running without any downtime. The switch to Curie was what resulted in the inappropriate text being generated. We leverage OpenAI’s content moderation tools, which have worked thus far for the Davinci model, but were not successful with Curie. We’ve been able to identify the root cause of our issue with the Davinci model, and will not be using Curie as a fallback in the future. We hope this sheds a little light on how this happened.
The experiment was going well until they ran into Twitch's guidelines policy, which caused the temporary suspension of the first A.I.-generated series.
