YouTube and Video
is launching a new feature called Corrections
, which lets you add prominent time-stamped corrections to existing videos in the description. This should be available at the end of June. Learn more from Creator Insider.
Webmasters and Publishers
Google is simplifying Search Console reports
by categorizing URLs as “valid” or “invalid” for a feature (like rich results). They will not show the current “warning” status when it might be confusing.
Google will be implementing new ad serving protections for teens
on August 15. Teens who are younger than 18 will not be shown personalized ads or ads from certain restricted categories. Those categories are not just alcohol, gambling and weapons, but also dating, body modification, contests and more. AdSense Publishers
should also be aware of any local restrictions on serving ads to teens.
There is a new “Google for Publishers” site
featuring AdSense publisher success stories. You might be featured there too. All you need to do is tell Google “how ad revenue has helped you grow your business”. Fill out the form here.
It’s the older folks Tweeting about US politics.
According to a recent Pew Research study,
it’s those age 50+ who produce about 50% of tweets from US adults, and nearly 80% of political Tweets. And here I thought the old folks were all on Facebook! (I kid. I’m an old.)
Alex Heath at The Verge has the scoop on an internal Facebook memo
that suggests making changes that would make the platform more TikTok
-like. The plan would be to show fewer posts from friends, family and accounts you follow in your feed, and more posts from “the best content that caters to your interests”. And more video Reels, of course. Casey Newton points out
that Facebook needs this pivot to a TikTok-like feed to work in order to fund their bigger pivot to the “metaverse”.
finally shut down completely this week
. Talk was Google’s original instant messaging platform which was migrated to Hangouts in 2017. But since then Talk could still be used via third party XMPP clients like Pidgin.
Does it think?
Experts say that there isn’t any evidence that LaMDA is anything more than a really sophisticated chatbot. And while Lemoine cites a conversation
where LaMDA stated it considers itself a person, in response to a question from the Post’s reporter, the AI said “I think of myself as an AI-powered dialog agent.” A previous demo had the AI chatting as Pluto
It seems that LaMDA very good at making statements that people want to hear. And there is concern
that the hype around whether the computer is “sentient” distracts from more important concerns about the current uses of AI.