在最近的一次讨论中，摩根士丹利的亚当·乔纳斯（Adam Jonas）强调了“实体AI”的经济影响，他预测，在埃德（Ed）和他退休之前，AI和机器人技术的结合将带来“8倍或10倍的全球GDP”乘数效应。他指出，截至2025年，全球GDP目前“大约为180万亿美元”，尽管此前的10倍增长花费了66年或150年，但由于“具身现实世界AI”的出现，下一次增长可能在短短30年内发生。 乔纳斯预计“未来一二十年内，数百亿机器人将进入我们的日常生活”。他指出特斯拉（Tesla）和SpaceX是这一转型中的关键公司，并表示它们“各自可以说已经构建了独立但协同关联的实体AI能力，并正在努力逐步弥合两者之间的差距”。他拒绝评论合并一事，但他认为它们之间的关系是“确定性的”，特斯拉是机器人的“制造和数据收集层”，SpaceX则提供“连接性和AI层”，两者都秉持着“大规模将能量转化为智能”的共同使命。 他断言“任何可以实现自动化的机器都将实现自动化”，并且任何没有AI边缘推理计算机的机器都将迅速过时。关于人形机器人，乔纳斯提到，摩根士丹利2050年的模型预测，在许多情景下，人形机器人数量将达到“数十亿”，尽管它们目前只充当“招聘工具和融资工具”。他设想未来将有“数千种不同层级的低空机器人、地面机器人、自动驾驶汽车、移动机器人、工业机器人以及介于两者之间的所有机器人”，它们以“群体智能”的方式连接，形成“连接到天基AI基础设施分布式轨道云的分布式推理云”。他将使用数据中心进行推理比作“开法拉利去买牛奶”，强调需要更高效的“边缘推理”。 乔纳斯认为特斯拉以及特斯拉/SpaceX提供了“最好的机会之一，以帮助西方工业复兴，并有机会在机器人技术和实体AI领域，随着时间的推移，赶上甚至超越我们的地缘政治竞争对手”。他宣称“自动驾驶汽车问题已解决”，指出“2023年Waymo从凤凰城撤走司机”是历史性的转折点，他相信这将“为其他形式因素的跟进打开闸门”。 播客主持人指出亚当·乔纳斯在特斯拉/SpaceX合并问题上“世界级的避而不答”，暗示这可能表明他掌握了特权信息。他同意乔纳斯关于AI的普遍性、“对智能近乎无限的需求”，以及因此“未来几十年对AI算力近乎无限的需求”的看法。播客主持人还强调了有能力扩展AI算力并发射“数百万颗用于推理的天基AI卫星”的公司的潜力。他最后提到，他新的“特斯拉和SpaceX 20年估值模型”现已在Patreon上发布，并推广AG1作为他日常健康方案的一部分。

In a recent discussion, Adam Jonas of Morgan Stanley highlighted the economic impact of "physical AI," predicting that the combination of AI and robotics will lead to an "8x or 10x global GDP" multiplier before Ed and he retire. He noted that global GDP currently stands at "$180 trillion ballpark as of 2025," and while previous 10x increases took 66 or 150 years, the next could happen in roughly 30 years due to "embodied real world AI." Jonas expects "tens of billions of robots in our daily lives within the next one to two decades." He pointed to Tesla and SpaceX as key companies in this transition, stating they "have each arguably assembled separate but synergistically linked physical AI capabilities and are working to gradually close the gap between them." While declining to comment on a merger, he believes their relationship is "deterministic," with Tesla acting as a "manufacturing and data collection layer" for robots and SpaceX providing "connectivity and the AI layer," both sharing a mission of "converting energy to intelligence at scale." He asserted that "any machine that can be automated will be," and any machine without an AI edge inference computer will face rapid obsolescence. Regarding humanoid robots, Jonas mentioned that in Morgan Stanley's modeling out to 2050, humanoids are projected to number in the "billions" in many scenarios, despite their current role as a "recruiting tool and a capital raising tool." He envisions a future with "thousands of different stratum of low altitude robots, terrestrial robots, autonomous vehicles, mobile robots, industrial, and everything in between," connected in a "swarming intelligence" for "distributed inference cloud connected to the distributed orbital cloud of the space-based AI infrastructure." He likened using a data center for inference to "using a Ferrari to pick up the milk," emphasizing the need for more efficient "inference at the edge." Jonas credited Tesla and Tesla/SpaceX with offering "one of the best chances to help revive and give Western industry a chance to keep up, if not to surpass our geopolitical rivals over time in robotics and physical AI." He declared that "autonomous cars are solved," pointing to "2023 when Waymo pulled the driver from Phoenix" as the historical turning point, which he believes will "open up the floodgates for other form factors to follow." The podcaster noted Adam Jonas's "world-class question dodge" regarding a Tesla/SpaceX merger, suggesting it might indicate privileged information. He agreed with Jonas on the ubiquity of AI, the "near-infinite demand for intelligence," and consequently, "near-infinite demand for AI compute for many decades to come." The podcaster also highlighted the potential for companies capable of scaling AI compute and launching "millions of space-based AI satellites for inference." He concluded by mentioning his new "20-year Tesla and SpaceX valuation model" is now live on Patreon and promoted AG1 as part of his daily health protocol.