When you buy through affiliate links in our content, we may earn a commission at no extra cost to you. Learn 
LinkedIn has begun testing a “Seems like AI slop” reporting option that lets users flag generic, repetitive, or automated posts they see in their feeds.
The company will use these reports to help it refine its models and improve what users see in their feeds. It’s also testing private warnings for users who’ve posted content that LinkedIn members have flagged as AI slop.
To encourage users to write in their own voice, the professional networking site is removing a feature that lets users rewrite posts with AI, replacing it with a more limited proofreading tool.
Before the change, the platform had given users the option of using an AI tool to polish their posts.
LinkedIn’s new approach will force it to determine when AI is used to help someone express an original idea, vs. when it’s simply used to produce generic content.
Asking its users to participate in that process could help improve the feed. However, there’s always the risk that users will start reporting carefully edited human writing or opinions they don’t like as AI slop.
LinkedIn Is Asking Users to Identify AI Slop
LinkedIn Chief Product Officer Hari Srinivasan announced the changes in a July 30 post, explaining they’re designed to reduce what the company has characterized as AI slop.
Once they’ve reported a post, LinkedIn removes it from their feed and uses the report as feedback to improve its recommendations.
Along with the AI slop button, the company will be using new classifiers to help it “identify if a post is AI-slop or generally low-quality content.” In the beginning, the classifiers will focus on suggested posts and content from outside a user’s network, since LinkedIn has more control over the content it recommends.
The networking site said it won’t treat all AI-assisted writing the same and acknowledged that “AI and slop are not the same thing.” In his post, Srinivasan pointed out that some people use AI because it makes them feel more confident about what they post.
Instead of relying on AI detectors to identify AI slop, the company says it wants feedback from people because automated tools aren’t infallible.
What Counts as AI Slop is Subjective
While the changes will likely be welcomed by people tired of scrolling through a LinkedIn feed full of AI-generated posts, they also create a new problem. People will be judging posts subjectively based on how they sound or look to them.
A human writer who uses the now-infamous em dash could see their post flagged as AI slop solely because they used a once-normal bit of punctuation that people now believe to be an indicator of AI-generated writing. The same thing could happen to a non-native English speaker whose posts are original but formulaic.
With a reporting button on every post, casual LinkedIn users are now empowered to accuse anyone of using AI, whether they actually did or not.
LinkedIn hasn’t said how many AI slop reports it would take before a user receives a private warning or whether these reports will affect a post’s reach. It also hasn’t said whether users accused of posting AI slop will be able to challenge the warning if they believe they’ve been wrongly flagged.
Removing its own AI rewriting feature may be the most effective way to discourage users from using AI to write their posts. Instead of helping users produce AI-edited posts that all sound alike, the platform will limit its assistance to proofreading. It hopes this change will help users write in their own voice.
The proofreading tool could do more to cut back on repetitive content than a reporting button that depends on users making subjective judgments.
LinkedIn is right: the feed needs to be cleaned up. However, it shouldn’t do so at the expense of penalizing people just because their writing looks or sounds like what others have come to associate with AI.
Also in Consumer Tech News
Apple May Charge Heavy Siri Users
Apple CEO Tim Cook has put Siri users on notice that if they use the AI assistant too much, they may at some point have to pay for the privilege.
Cook made those statements during Apple’s July 30 earnings call. He said the company expects some people to use its new AI-powered version of Siri frequently and plans to offer upgrade options through iCloud+.
The tech giant is still working out how it’ll cover the additional computing costs, so it hasn’t yet announced prices, usage limits, or subscription tiers.
Apple hasn’t said how much Siri AI access will come standard with its devices. Cook only said that users could have the option to move up to a higher iCloud plan, but the details haven’t been made public.
The CEO’s comments suggest that buying an AI-enabled iPhone may not include unlimited access to all of the assistant’s capabilities.
Roku Streaming Devices Get Much More Expensive
Roku has hiked prices across its streaming device lineup, with some models now costing as much as 60% more.
The standard Streaming Stick went from $29.99 to $39.99, while the Streaming Stick Plus increased from $39.99 to $59.99. The Streaming Stick 4K jumped from $49.99 to $79.99, and the Roku Ultra rose from $99.99 to $149.99.
In a July 24 call with The Desk, a Roku executive blamed the ongoing memory shortage, which keeps getting worse due to the AI boom.
Consumers have traditionally looked at streaming sticks as a cheap way to upgrade their TVs. But at $80 a pop, the Streaming Stick 4K is entering a price range where people might start comparing it with more powerful devices or just use the streaming software already installed on their TVs.
Spotify Turns Playlists Into Running Sessions
Spotify announced on July 30 that it had launched Running Mode. The feature allows Premium users to create personalized running sessions based on their workout and music preferences.
Users can choose from 25 presets and customize the workout type, duration, beats per minute, and music. Running Mode also includes optional English-language audio cues for guidance and motivation.
The new mode is only available to iOS users in the United States, Canada, the United Kingdom, Ireland, Australia, New Zealand, and Sweden through Spotify’s Fitness Hub.
By introducing this feature, Spotify is moving closer to fitness services like Peloton and Apple Fitness+, while also focusing on its biggest strength: matching music to a listener’s activity and preferences.
DoorDash Is Building Its Own Delivery Drones
In a July 29 statement posted to its website, DoorDash said it has received approval from the Federal Aviation Administration to operate its own commercial drone-delivery service.
The company earned a Part 135 air carrier certificate and launched DoorDash Air, an in-house program that develops drones, merchant handoff systems, safety operations, and delivery software. According to DoorDash, it’s the eighth drone operator to receive the certification.
The drone setup will eventually allow its platform to decide whether an order should go to a human Dasher, a ground robot, one of its partners, or a DoorDash drone. It should speed up the delivery process as well.
DoorDash hasn’t said when customers will start receiving drone deliveries. It plans to reveal more about the drone and its rollout later this year.
