Can anyone help me identity this mushroom? It grew up in just a few days in my garden, and it’s currently over a foot wide. I’m definitely not going to eat it, I’m just curious!

There’s nothing ‘unofficial’ about this start to the summer.

The stars of the Memorial Day Fair, the town farm’s goats!

First, asymmetry of attention: We know some posts get a lot of views/engagement and others don't, but how wide is the gap? We used the Gini Index to measure. For reference, income inequality in the US come in at 41.6. Engagement inequality on Facebook is .92! 1% of posts got 58% of all engagement.

I feel so bad for the physicists. In six weeks they'll be begging in the streets to fund their supercolliders. Whereas over here in the computer science faculty we can have a robot sale that will keep the compute cluster going for MONTHS.

So proud of @northeasternu.bsky.social ‘s Tech Policy team! Kate, Michelle, Fernando, and Jayant participated in the @atlanticcouncil.bsky.social ‘s Cyber 9/12 Policy Competition and did brilliantly!

Chiefly, we wanted this system to be useful to users. So, we also build a visualization: a 'Feed Card' that summarizes all the key characteristics in one image. Here's the one for X: 9/12

Content Timeliness: How heavily recent content is prioritized over older material. Some types of content are evergreen (art history, for example), while others (news about a breaking event) are only relevant if they are very recent. 7/12

Novelty: How much the algorithm introduces unfamiliar content versus reinforcing known preferences. Algorithm designers mostly want to show users content they like, but algorithms sometimes introduce novel content that they have no way to know how a user will react to. 6/12

Specificity: Whether content is targeted narrowly to individual interests or broadly to maximize audience size. Very broadly appealing content that many people like also tends to be content that very few people love or feel strongly about. 5/12

Usage Intensity: Whether the platform optimizes for active engagement (commenting, sharing) or passive consumption (watching, browsing). When people have intense interactions and experiences, whether they are positive or negative, they tend to end them sooner. 4/12

New work by me from @knightcolumbia : Into the Driver's Seat with Algorithmic Feeds. Wonder how, exactly, how social media algorithmic feeds work? They're incredibly opaque, so I've spent years building a framework for measuring and explaining them. Buckle up and come with me!🧵

As you likely know, this isn’t caused by the legal issue - it’s a business decision, triangulating around political realities. But believe me when I tell you that there are so many times platforms pull stunts like this, blaming a technical issue when what’s going on is brinkmanship.

Now you might ask, is he talking about predicting what user will like based on their own explicit signals or are they using them as inputs into the recommender system? And I read these sentences that mention prediction and recommendation to mean 'both'.

First of all, let me say that I think it's ridiculous that we have to parse Meta's blog posts like Kreminologists. BUT: I'm reading this statement of 'ranking and showing you content based on explicit signals' of which reshares is one.

Mark's statement had 3 main points. I'm going to talk about the last one.

This is a nice DSA you've built here, *cracks knuckles*. It would be a real shame if my friend Pres. Trump got upset and something happened to it.
www.techpolicy.press/transcript-m...

Great start to the World Championships @nhrl.bsky.social !

In personal news, as they say… I’m starting as an Assistant Professor of Computer Science at Northeastern University next week. I’ll be recruiting PhD students this year, so if you want to work with me you should apply to KhouryCollege!