发布 HN：Societies.io（YC W25）– 针对目标受众的 AI 模拟
嗨，HN，我们是Patrick和James！人工社会（<a href="https://societies.io">https://societies.io</a>）让您能够模拟目标受众，以便在发布营销、信息和内容之前进行测试。
<p>这是一个快速的产品演示：<a href="https://www.loom.com/share/c0ce8ab860c044c586c13a24b6c9b391?sid=b7b4e23c-5223-4aad-ae08-b5c704d45aad" rel="nofollow">https://www.loom.com/share/c0ce8ab860c044c586c13a24b6c9b391?...</a>
<p>市场营销人员总是说他们的一半支出会被浪费——他们只是不知道是哪一半。现实世界的实验有帮助，但它们在大规模运行时太慢且成本太高。因此，我们正在构建模拟，让您能够快速且低成本地测试，以找到最佳的信息版本。
<p>工作原理：
<p>- 我们根据来自实际个人的真实数据创建AI角色，这些数据来自公开的社交媒体资料和网络来源。
<p>- 对于每个受众，我们从数据库中检索相关角色，并在一个互动的社交网络图上进行映射，旨在复制社会影响的模式。
<p>- 一旦您草拟了信息，每个实验会运行一个多智能体模拟，角色会对您的内容做出反应并相互互动——这些模拟运行时间为30秒到2分钟。然后，我们会提供结果和见解，帮助您改进信息传递。
<p>我们面临的两个最大挑战是准确性和用户界面。我们已经测试了我们在预测LinkedIn帖子表现方面的性能，初步结果令人鼓舞。我们的模型R²值为0.78，我们发现“信息传播”在我们的模拟中是预测实际互动的最重要因素，尤其是当查看同一作者发布的帖子时。但在将这些模拟推广到其他背景以及找到评估的真实数据方面，还有很长的路要走。关于准确性，我们这里有更多信息：<a href="https://societies.io/#accuracy">https://societies.io/#accuracy</a>
<p>在用户界面方面，我们面临的最大挑战是确定“实验”这种形式是否对用户有吸引力。我们故意将重点放在这一点上（而不是AI调查），因为实验利用了我们在社会影响和想法如何在角色之间传播方面的专业知识。
<p>James和我都是行为科学家，但走了不同的道路来到这里。我帮助企业进行A/B测试，以提升销售和客户留存率。同时，James成为了一名数据科学家，并在业余时间将33,000个大型语言模型聊天机器人连接在一起，并撰写了一篇相关论文（<a href="https://bpspsychub.onlinelibrary.wiley.com/doi/pdfdirect/10.1111/bjop.12764" rel="nofollow">https://bpspsychub.onlinelibrary.wiley.com/doi/pdfdirect/10....</a>）。他向我展示了这些模拟，我们决定将其做成一家初创公司。
<p>定价：人工社会可以免费试用。新用户可获得3个免费积分，然后享受两周的免费试用。专业账户每月40美元可获得无限模拟。我们计划稍后推出团队功能，以及为定制受众提供企业定价。
<p>我们希望您能试用这个工具并分享您的想法！
查看原文
Hi HN, we’re Patrick and James! Artificial Societies (<a href="https://societies.io">https://societies.io</a>) lets you simulate your target audience so you can test marketing, messaging and content before you launch them.<p>Here’s a quick product demo: <a href="https://www.loom.com/share/c0ce8ab860c044c586c13a24b6c9b391?sid=b7b4e23c-5223-4aad-ae08-b5c704d45aad" rel="nofollow">https://www.loom.com/share/c0ce8ab860c044c586c13a24b6c9b391?...</a><p>Marketers always say that half their spend will be wasted - they just don’t know which half. Real-world experiments help, but they’re too slow and expensive to run at scale. So, we’re building simulations that let you test rapidly and cheaply to find the best version of your message.<p>How it works:<p>- We create AI personas based on real-world data from actual individuals, collected from publicly available social media profiles and web sources.<p>- For each audience, we retrieve relevant personas from our database and map them out on an interactive social network graph, which is designed to replicate patterns of social influence.<p>- Once you’ve drafted your message, each experiment runs a multi-agent simulation where the personas react to your content and interact with each other - these take 30s to 2 minutes to run. Then, we then surface results and insights to help you improve your messaging.<p>Our two biggest challenges are accuracy and UI. We’ve tested our performance at predicting how LinkedIn posts perform, and the initial results have been promising. Our model has an R2 of 0.78 and we’ve found that “message spread” in our simulations is the single most important predictor of actual engagements when looking at posts made by the same authors. But there’s a long way to go in generalising these simulations to other contexts, and finding ground truth data for evals. We have some more info on accuracy here: <a href="https://societies.io/#accuracy">https://societies.io/#accuracy</a><p>In terms of UI, our biggest challenge is figuring out whether the ‘experiment’ form factor is attractive to users. We’ve deliberately focused on this (over AI surveys) as experiments leverage our expertise in social influence and how ideas spread between personas.<p>James and I are both behavioral scientists by training but took different paths to get here. I helped businesses run A/B tests to boost sales and retention. Meanwhile, James became a data scientist and, in his spare time, hooked together 33,000 LLM chatbots and wrote a paper about it (<a href="https://bpspsychub.onlinelibrary.wiley.com/doi/pdfdirect/10.1111/bjop.12764" rel="nofollow">https://bpspsychub.onlinelibrary.wiley.com/doi/pdfdirect/10....</a>). He showed me the simulations and we decided to make a startup out of it.<p>Pricing: Artificial Societies is free to try. New users get 3 free credits and then a two week free trial. Pro accounts get unlimited simulations for $40/month. We’re planning on introducing teams later, and enterprise pricing for custom-built audiences.<p>We’d love you to give the tool a try and share your thoughts!