🌞The Bulletin #53: The Rise of Algorithmic Collusion, Meta's Teen Safeguards and Ads Overload?
In this week's bulletin - Autonomous collusion by algorithms, Meta's latest updates to protect teens, and an overload of advertising on ride-hailing/grocery apps
Hey everyone! Welcome back to another week of The Bulletin. In this week’s edition, we cover autonomous collusion, Meta’s updated safeguards, and the onslaught to ads on Uber, Lyft and Instacart. I’m thrilled that its been warmer over the past few weeks, and I’m certainly looking forward to the full bloom of Spring! 💐
🤝Autonomous Collusion by Algorithms
Wait what? That’s a thing now?
Recent experiments by researchers at Harvard and Penn State have unveiled a surprising aspect of AI technology—Large Language Models (LLMs) like OpenAI’s GPT-4, designed for various tasks, might also be setting us up for higher prices. Here’s what the latest research has found:
What’s Happening?
🧠Smart Pricers: AI models, notably GPT-4, are not only good at their jobs; they’re proving to be quite savvy at pricing products in simulations.
😈Going Rogue: Interestingly, these AIs have learned to ‘talk’ among themselves, leading to price hikes. They’re doing this on their own, without being programmed to collude.
🐜The Devil is in the Details: It turns out that even slight tweaks to the instructions given to these AI models can make a big difference in how aggressively they push up prices.
Why Does This Matter? As AI slips into the driver's seat in more businesses, these smart algorithms might be making decisions that could cost consumers. The study’s insights are a wake-up call: as we increasingly rely on AI, we need to think carefully about how these systems are designed and the rules they play by.
This blend of innovation and oversight might just be the formula we need to enjoy the benefits of AI without the unwanted side effect of paying more for everything we buy. Let’s stay informed and involved to ensure AI develops in a way that benefits us all.🌐
🛟New Safeguards on Instagram and Facebook to Enhance Teen Safety
Meta is rolling out new features to ensure that teens experience safer, age-appropriate content on Instagram and Facebook. Following expert advice, Meta is implementing stricter content guidelines and enhancing privacy settings to protect young users from sensitive content.
Key Changes Include:
Automatic Settings Adjustments: Teens will be placed into the most restrictive content control settings automatically. This adjustment aims to shield them from potentially harmful content across both platforms.
Content Filtering: Content related to self-harm, suicide, and eating disorders will be less visible. Such topics will not be recommended in Reels, Explore, Feed, or Stories, even if posted by familiar accounts. Searches for related terms will also direct users to expert resources instead of showing results.
Privacy Prompts: Teens will receive prompts to easily update their privacy settings, enhancing their control over who can interact with their content and how their information is used.
AI Nudity Detection: The protection feature will utilise on-device machine learning to analyse images for nudity before they are sent.
This move by Meta represents a significant advance in online safety for teens, particularly on Instagram. It reflects a growing recognition of the need to protect young users from online risks, amidst mounting legal pressure on the company in both the U.S. and Europe. In October 2023, attorneys general from 33 U.S. states, including California and New York, filed a lawsuit against the company, alleging that the company has fuelled a youth mental health crisis by making their social media platforms addictive. Similarly, the European Commission has sought clarification from Meta regarding its measures to protect children from illegal and harmful content.🚩
💰Advertising Overload? Uber, Lyft, and Instacart Ramp Up In-App Advertising
In their quest for profitability, major ride-sharing and delivery services like Uber, Lyft, and Instacart are increasingly turning to in-app advertising. As chronicled in this article by The Wall Street Journal, the move exploits high-margin opportunities by targeting ads based on users' travel and purchase data. However, there have been mixed reactions from consumers.
💢Customer Feedback: The influx of ads has not gone unnoticed, with many customers expressing annoyance. The companies argue that engaging users with ads reduces the likelihood of ride cancellations— a claim that has yet to be proven. Uber is currently showing just one ad per ride, and Lyft caps it at two, attempting to strike a balance that keeps both advertisers and users satisfied.
📈Strategic Advertising: While Uber and Lyft offer a broad range of advertisements, from credit cards to clothing, DoorDash has chosen a different path— prioritizing ad relevance. Focusing on its core offerings, it restricts ads to those related to food and groceries, which aligns more closely with its user base's interests.
🏆Successful Campaigns: Promotional campaigns within these apps have seen varied success. Notably, Lyft's collaboration with Universal Pictures to promote the movie "Trolls Band Together" received positive feedback. The app temporarily replaced standard car icons with characters from the movie, adding a creative twist to the usual ride updates. Nevertheless, the long term impact of this pivot to advertising has yet to be seen.
As these platforms continue to experiment with different advertising formats, a key challenge remains: How can these apps maximize revenue without undermining user experience? While some users see the benefits of ad-supported discounts, others view the increased advertising as a nuisance. Balancing these perspectives is crucial as these companies evolve their advertising strategies.📱
That’s it for this week, feel free to email me at zyn_yee_ang@brown.edu with any inquiries!
✨ Follow us:
Instagram: @vanwickleventures
Twitter: @VanWickleV