这段来自我的@warpsciences联合创始人Sol Reisberg的观点非常重要——这是一位行业专家的精彩阅读材料。https://www.linkedin.com/posts/sol...
This take from my @warpsciences co-founder Sol Reisberg is important—and a great read from an expert in the space. https://www.linkedin.com/posts/sol...
Kevin WeilAI2026-10-24
所罗门·赖斯伯格 2天前 发布此帖 今天我们宣布了Warp Sciences。凯文·威尔已经写下了他的愿景，虽然我赞同它，但我还想从实验室的角度补充一下，因为我是创始团队中的化学家。说实话，我对此持怀疑态度。我花了整个职业生涯来建立公司以发现药物，并见证了“人工智能将改变药物发现”这一承诺一次又一次地被提出。但它从未实现过。大多数声称这一点的公司从未真正合成过化合物，从未运行过检测，也从未向项目负责人解释过为什么化学研究推迟了一个月。他们是在卖一个没有走过这条路的目的地。我在凯文·威尔、DJ帕蒂尔和马诺哈尔·帕鲁里身上找到了相反的情况。没有人是在卖一个解决科学问题的方案。我们都在围绕同一个狭窄而困难的问题打转，讨论的是为什么它可能不会成功。这种脚踏实地和谦逊在这个科技领域是罕见的，这也是为什么我想和他们一起建立Warp的原因。这个问题既不耀眼又真实。无论你是否喜欢，大多数西方生物技术公司的临床前化学研究都是在中国的实验室完成的，价格是美国实验室第一天就无法比拟的：一个先导化合物在海外大约需要1000美元，而在美国做同样的工作，在任何人拿起烧瓶之前，就已经包括了房地产、试剂和溶剂，这些成本比美国高出60-80%。客户愿意为国内工作支付溢价，但这只是一个溢价，而不是一个倍数。人工智能即将使化合物设计变得便宜，这意味着合成和测试它们成为瓶颈。这就是我们的业务：一个下一代合同研究组织（CRO），为真正的客户提供真正的化学服务，并一次自动化一个工作流程，直到经济可行。这是一条艰难的道路，我们直言不讳。这里有一个墓地：许多团体都试图大规模自动化化学并失败了，更多的人在研究了经济因素后决定不尝试。仅仅供应链（试剂、溶剂、构建块，这些在美国的价格点还没有人触及）就是一个需要多年的问题。利润率将...
原文
Solomon Reisberg 2d Report this post Today we’ve announced Warp Sciences. Kevin Weil has written about the vision, and while I share it, I want to add the view from the bench, because I’m the chemist on the founding team. Honestly, I came to this with a lot of skepticism. I’ve spent my career building companies to discover drugs, and watched “AI will transform drug discovery” get promised again and again. It still has never delivered. Most companies making that claim never had to make a compound, run an assay, or explain to a program lead why the chemistry slipped a month. They were selling a destination without having walked the road. What I found in Kevin Weil, DJ Patil, and Manohar Paluri was the opposite. Nobody was selling a solve for science. We were all circling the same narrow, hard problem, and the conversations were about why it might not work. That groundedness and humility is rare in this corner of tech, and it’s why I wanted to build Warp with them. The problem is unglamorous and real. Like it or not, most preclinical chemistry for Western biotech is done in China, at a price a US lab can’t match on day one: a lead-opt compound runs about $1,000 offshore, and the same work in the US starts with real estate, reagents, and solvents that cost 60-80% more, before anyone even picks up a flask. Customers will pay a premium for domestic work, but it’s a premium, not a multiple. AI is about to make designing compounds cheap, which means making and testing them becomes the bottleneck. That’s the business: a next-generation CRO that does real chemistry for real customers, and automates it one workflow at a time until the economics work. That’s a hard road, and we say so. There’s a graveyard here: plenty of groups have tried to automate chemistry at scale and failed, and more have looked at the economics and decided not to try. The supply chain alone (reagents, solvents, building blocks at US price points nobody has hit yet) is a multi-year problem. Margins will be thin for a while. None of this can be solved by a model or a pitch deck. Can we build the automation? It's a real risk, and one this team is built to take on. But the deeper question: even if the robots work, is it worth it? I think it is, for three reasons. 1. The market grows as AI floods the funnel with ideas that have to be made. 2. Costs fall with scale, and no one in the US has yet scaled to where that pays off. And 3. The data an automated lab generates is worth more than the compounds it makes. We’re not claiming we’ll 'cure all disease,' or that we'll cure even one; our customers will do that. We’re claiming we can bring a critical piece of drug discovery back to the US, make it faster and cheaper over time, and build something durable along the way. It'll be very hard, but it's not fantasy. If you’ve ever wanted to be the person actually making the molecule, come build this with us. https://warp.xyz/careers/ Kevin Weil Kevin Weil is an Influencer CEO at Warp Sciences 2d 💥