Due to the continuous use and rise of social media sites among teens and worries about their safety while using these sites, Human-Computer Interaction (HCI) researchers have been working together with teenagers using methods such as participatory design/co-design to study their social media usage, along with the risks they encounter and their behavior online. However, this comes with several constraints such as power imbalances, conflicting schedules of teenagers, and lack of attendance. To work towards engaging teens in online safety research without being bound by these constraints, we recruited 7 teens between the ages of 15-17 years old to be a part of a year-long Youth Advisory Board (YAB) for the Socio-Technical Interaction Research (STIR) Lab. We conducted online activities with the teen participants in our YAB within an Asynchronous Remote Community (ARC), which is an online group where communication and activity participation occur asynchronously between researchers and participants. To collect our data, we first organized a synchronous Zoom meeting with the teens about the ARC research method, discussed how the method relates to them, and talked about the topics and activities that they would like to partake in within an ARC. Then, we conducted activities with the teens to discuss their social media usage and their perceptions of algorithms used to provide content on social media. Our participants were interested in discussing topics that were related to safety on social media platforms and doing research activities that could best be implemented through the ARC research method such as giving feedback on research design, suggesting solutions to online safety prompts, and designing apps based on an online safety prompt. We also found that our participants' social media platform usage heavily depended on their goals as users, and they were weighing the privacy affordances and risks that came with using the platforms to decide which platform is best for their goals. Finally, our teen participants were aware of the positive and negative impacts that algorithms on social media platforms have on their online safety. They were skeptical of social platform algorithms due to their safety and privacy concerns, though they understood the benefits and wanted to learn more about how these algorithms operate. Overall, our participants appreciated the ARC methodology for exploring online safety perspectives, as it does allow them enough time to take part in activities and collaborate in discussions with each other. Our findings suggest that ARC can be an effective method for co-designing research studies to understand teens' social media usage, the risks they encounter online, and ways to mitigate such risks on social media.