The Morning After: Midjourney shutters free trials of its AI image generator due to ‘extraordinary’ abuse
It’s a day of reality catching up with the chatbot boom. In the last 24 hours alone, we’ve had hoaxes, FTC complaints and… ads. Hooray. We’ll get into how Microsoft is bringing ads to its Bing chatbot – bound to happen – while OpenAI may have to halt ChatGPT releases in the face of FTC complaints.
The nonprofit research organization, Center for AI and Digital Policy (CAIDP), says OpenAI’s models are “biased, deceptive” and threaten privacy and public safety. The CAIDP says OpenAI also fails to meet Commission guidelines calling for AI to be transparent, fair and easy to explain. There’s no guarantee the FTC will act on the complaint. If it does set requirements, though, the move would affect development across the AI industry.
– Mat Smith
The Morning After isn’t just a newsletter – it’s also a daily podcast. Get our daily audio briefings, Monday through Friday, by subscribing right here.
The biggest stories you might have missed
Uber adds 14 new cities to its EV rideshare service
‘Star Trek: Picard’ embraces its nihilism
Apple’s ‘Tetris’ movie trades real-life drama for spy fantasies
Github ordered to identify user who leaked Twitter source code
Sam Bankman-Fried pleads not guilty to latest fraud, bribery charges
Microsoft explains how ads will happen in Bing’s AI chatbot
Ah, the end of the fun.
Over the past few days, users have reported seeing ads inside Microsoft’s Bing chatbot experience. Based on the limited examples we’ve seen, the GPT-4-powered chatbot embeds relevant ad links in response to users’ actual questions. Ads don’t seem to appear for most people (including us) yet, but they’ll most likely pop up more frequently and in more places soon. In a post on the Bing blog, Microsoft Corporate VP for Search and Devices Yusuf Mehd, explained that ads would come in the form of a linked citation, along with additional links in a “Learn More” section below Bing’s response to their query. In the future, Microsoft could add functionality where hovering over a link from an advertiser would display more links from its website to drive more traffic to it.
Midjourney ends free trials of its AI image generator due to ‘extraordinary’ abuse
The tool had been used to fake images of Trump and the Pope, among others.
Midjourney CEO, David Holz, announced on Discord that the company is ending free trials due to “extraordinary demand and trial abuse.” New safeguards haven’t been “sufficient,” and you’ll have to pay at least $10 per month to use the image generator going forward. As The Washington Post reported, Midjourney has picked up unwanted attention in recent weeks. Users relied on the company’s AI to build deepfakes of Donald Trump being arrested, and Pope Francis wearing a trendy coat.
Polestar 3 first look
Possibly the best-looking EV of the year.
The Polestar 3 was recently showcased in New York for its North American debut, so we had to check it out. It might just be the best-looking new SUV in 2023. The Polestar 3 is built on the same platform as the Volvo EX90, but the company has made some significant changes that ensure there won’t be confusion between the two. Instead of three rows of seats, the Polestar 3 maxes out at two, with slightly less rear storage in favor of a more spacious cabin.
A new Twitter clone is trying to seduce original blue checkmark owners
T2 is led by former Twitter employees who want to recreate Twitter’s “public square.”
With “legacy” Twitter checkmarks about to disappear (tomorrow!), one Twitter alternative hopes to lure some of those OG verified users to its platform. T2, an invite-only service led by two former Twitter employees, says it will allow accounts to carry over their “legacy” Twitter verification to its site. T2 is part of a growing crop of Twitter alternatives that have sprung up after Musk’s takeover. Founder Gabor Cselle has been clear that he intends to create “a pretty straightforward copy of Twitter with some simplifications”.
Netflix is testing TV games that use phones as controllers
Hidden code references games on TV in the Netflix app.
Netflix might have started (or is at least looking to start) testing games for TV, based on code within its app that developer Steve Moser shared with Bloomberg. Moser reportedly found hidden references to games played on television, as well as additional code that indicates the possibility of using phones as controllers to play them. One line from within the app apparently reads: “A game on your TV needs a controller to play. Do you want to use this phone as a game controller?” The streaming giant launched several games on Android, iPhones and iPads in 2021, but on the Netflix app for TV, these games were notably absent.
This article originally appeared on Engadget at https://www.engadget.com/the-morning-after-midjourney-shutters-free-trials-of-its-ai-image-generator-due-to-extraordinary-abuse-111518140.html?src=rss
It’s a day of reality catching up with the chatbot boom. In the last 24 hours alone, we’ve had hoaxes, FTC complaints and… ads. Hooray. We’ll get into how Microsoft is bringing ads to its Bing chatbot – bound to happen – while OpenAI may have to halt ChatGPT releases…
