Digital Digest #43 - ChatGPT image editing, Meta fights AI-fueled manipulation, LinkedIn launches CTV ads

Hi everyone,

Happy Monday! Hope you had a great weekend.

Welcome to this week's edition of Digital Digest! In this week's edition: X rolls out Grok to premium users, Meta begins to fight AI-fueled misinformation by labeling AI-generated content, and LinkedIn moves into Connected TV… among many other stories.

But first, ChatGPT has rolled out the ability to edit images created through its platform, using text prompts.

1. ChatGPT introduces image editing

OpenAI rolled out the ability to edit images created in ChatGPT, using text prompts. (Axios, Creative Bloq, Tom's Guide, Mashable)

• Why it matters: You know the feeling. Your image is almost right, but you just want one thing to be different. So you tell ChatGPT to tweak it… and get back something completely different. Sigh. Well, no more - now you can highlight where you want an edit to be made, and get a new version with your tweak incorporated. This isn't novel - other tools have had this feature for a while - but incorporating it into ChatGPT makes it even more available to the masses.

Also in generative AI:

• ChatGPT won’t require you to log in to use it anymore (Axios)
• How agencies use AI to generate new business (AdWeek - subscription required)
• How advertising agencies discuss AI in pitches (eMarketer)

2. X makes Grok available to premium users

X (Twitter) is rolling out access to its Grok AI platform for premium X subscribers (yes, both of them… snicker). (TechCrunch)

• Why it matters: With its access to real-time data from Twitter (sorry, X), Grok can do interesting things like summarizing the news. Unfortunately, with the hellscape that X has now become, those headlines have already been… sketchy.

Also on X:

• X announced a "significant initiative" to purge bots and spam accounts from the platform. (X, Social Media Today, TechCrunch)
• X owner/saboteur Elon Musk has continued to spread false and harmful misinformation on his platform. (Ars Technica, Gizmodo)

3. Meta updates AI labeling in its apps

Meta has updated its AI labeling policies and will add "Made with AI" labels to content when it detects "industry standard AI image indicators or when people disclose that they're uploading AI-generated content." (Meta, TechCrunch, Social Media Today, Ars Technica)

Why it matters: This seems minor, but it's a big deal in my view. I'm pretty thrilled with this update, which Meta is making in response to feedback from its oversight board. AI-generated content (and the growing threat of disinformation) are rapidly raising the risk of people being manipulated by what they see online. This move by Meta - one that I'm hopeful will include support for metadata-focused standards like Content Credentials - is an important step in protecting the broader online population from these harms.

Also in misinformation:

• Fake AI law firms are sending fake DMCA threats to generate fake SEO gains. (404 Media, Ars Technica)
• "Pink slime" local news outlets erupt all over US as election nears (Ars Technica)

4. LinkedIn launches Connected TV and live events ads

LinkedIn announced a slew of new ad types, including the ability to set up Connected TV ads from within the platform across publishers including Paramount, Roku and Samsung Ads, along with a partnership with NBCUniversal to enable ads across their streaming content. The platform also announced Live Event Ads to enable promotion of events before, during and after they take place. (LinkedIn, Social Media Today, AdExchanger)

• Why it matters: A way to reach business decision-makers in a targeted way via Connected TV? Drool. Live Event Ads also fill a gap, and the dynamic adjustment of the ad units based on timing that this feature enables is a nice touch. Also: LinkedIn appears to be rolling out generative AI engagement prompts in its news feed, beneath organic content. Clicking the prompts pops up a chat box with a response to the question, with a text box for you to enter more questions. I can't link to news because… I just spotted it in my feed but can't find stories about it. Look at me, breaking news…

Also on LinkedIn:

• New study looks at most effective LinkedIn post types in 2024 (Social Media Today)

Have a great week!

Dave