KGI's Peter Chapman was featured on @techpolicypress, breaking down AliExpress' recent binding commitments under the EU’s Digital Services Act on six key areas including recommender and advertising system transparency and researcher data access. Read here:

🚨NEW Report: The Technical Feasibility of Divesting Google Chrome 🚨 . Our report shows a Google Chrome divestiture is technically feasible and offers an engineering assessment of what it would take to separate Chrome from Google – and succeed #USvGoogle.

RT @lutherlowe: An interesting and important report by @knightgtown on the technical feasibility of a potential Chrome spinoff. Great resea….

@PinterestEng @Pinterest 🎯Our takeaway: Explicit user feedback can complement behavioral signals and help improve both relevance and satisfaction in recommender systems. Kudos to the @PinterestEng team for taking the lead here, and for sharing what they've learned. 👏👏

@PinterestEng @Pinterest It’s a shift from optimizing for what grabs attention to optimizing for what adds value – a crucial step toward healthier, more relevant personalization. /6.

@PinterestEng @Pinterest Online user survey data helps in 3 key ways:.✅Identifying content users like but don’t engage with.✅Reducing low-quality or frustrating recommendations.✅Re-ranking based on perceived value, not just behavior. /5.

@PinterestEng 🤔Why does this matter?.💡Because engagement can amplify polarizing, low quality and harmful content. By incorporating survey responses, @Pinterest can more accurately align with what users actually value. /4.

@PinterestEng .@PinterestEng started asking users directly: . ✅“How relevant or irrelevant is this content for you?” ✅"Does this content inspire you?”. This qualitative feedback is used to provide more relevant AND valuable content recommendations. /3.

🤳Some platforms optimize recommendations primarily by using engagement signals (clicks, saves, likes). But these don’t always lead to positive online experiences. @PinterestEng wanted to root out negative and harmful experiences, so. /2.

How can platforms build better algorithmic feeds? . @Pinterest @PinterestEng shares its approach: using user surveys to guide its recommender system – reducing harm and increasing relevance. Here’s how it works 👇.🧵.

RT @PinterestEng: New article on the Pinterest Engineering Blog 💻 Read more about how Pinner surveys help us move toward "healthier persona….

This position is based at @Georgetown in Washington, D.C., with a research focus on how competition authorities worldwide are evolving their strategies to address concentration in digital markets. Applications reviewed on a rolling-basis, with a position start date of Aug 2025.

🚨New exciting job opportunity! Applications are open for our Postdoctoral Fritz Fellow in Digital Competition Research! Read more below. Apply here today 👉

"The #Google #antitrust search case was always about AI. if you look at the backend of Google's Gemini, it directly benefits from all their user data and Google's search index. No competitor has those advantages. Data is so fundamental to all these AI products." @alissacooper.

"We will know these remedies have been successful when there is entry into the market" KGI's @alissacooper speaks at @GWlaw & @Crowell_Moring Antitrust & Tech conference on the recently litigated monopolization cases against Google + how remedies can be crafted most effectively.

RT @FloorCharts: .@alissacooper of @knightgtown emphasizes the need to think upstream about speech, platforms, and AI governance because:….

RT @Ashkhen: @RSI @SpenceKjell We should learn from the lessons of social media and how to design product for better goals and how to not d….

@alissacooper @SayreEvan The episode also focuses on why design-based approaches for creating better feeds are a more fertile ground to explore than content moderation, which can feel like an unsolvable problem. 🎙️Listen here:

@alissacooper and The Dynamist host @SayreEvan.dig deep into how social media feeds can be better designed to maximize long-term user value and shift away from attention-maximization designs that can lead to harms and negative online experiences.

🤔Is there a better approach to creating better social media algorithms than an outright ban on personalized feeds? Our Executive Director @alissacooper was featured on @JoinFAI's The Dynamist podcast to help answer this question. [thread]