RT Avi Chawla (@_avichawla)You're in a Research Scientist interview at OpenAI.The interviewer asks:"How would you expand the context length of an LLM from 2K to 128K tokens?"You: "I will fine-tune the model on longer docs with 128K context."Interview over.Here's what you missed: ...

You're in a Research Scientist interview at OpenAI.The interviewer asks:"How would you expand the context length of an LLM from 2K to 128K tokens?"You: "I will fine-tune the model on longer docs with 128K context."Interview over.Here's what you missed: ...

1) From the perspective of AI smartphone agents: Goldman Sachs does and seeks to do business with companies covered in its research reports. As a result, investors should be aware that the firm may have a conflict of interest that could affect the objectivity of this report. Investors should consider this report as only a single factor in making their investment decision. For Reg AC certification and other important disclosures, see the Disclosure Appendix, or go to www.gs.com/research/hedge.html. Analysts ...

本文来自微信公众号： 新智元 ，作者：新智元，编辑：好困、定慧，题图来自：AI生成 如果将时间拨回三年前的今天，也就是2022年12月1日，那是一个相对安静的周三。 位于旧金山的一家名为OpenAI的非营利实验室，悄无声息地发布了一个名为"ChatGPT"的研究预览 版。 没有盛大的发布会，没有乔布斯式的演讲，只有一个朴素的对话框。 当时的人们并不知道，这个对话框将彻底改变世界。 ChatGPT早已不是那个偶尔会算错数学题的聊天机器人，它和它的继承者、竞争者们已经成为了人类 在数字AI世界赖以生存的"氧气"。 然而，伴随着技术的指数级跃迁，一种难以名状的群体性焦虑正在全球蔓延，和每个人都息息相关。 ChatGPT三年前的样子 三年后的今天，2025年12月12日，当我们站在这个时间节点回望，世界已经被彻底重塑。 这三年里，围绕ChatGPT和生成式AI，我们见证了前所未有的狂热与恐慌交织在一起：硅谷高歌猛 进，华尔街亦疯狂逐利，但普通人和各行各业从业者却充满焦虑和不安。 正如《大西洋月刊》评论所言，我们正身处"ChatGPT建造的世界"！ 一个充满不稳定性的时代，大家都在战战兢兢地等待下一只靴子落地。 年轻人 ...

How do you start on a project like this. Like I think a thing that I see a lot of AI developers and people building stuff with these tools do wrong is they try to do the like 100% version of the thing all at once. To me, it seems like the struggle right now is how do we sort of sequence this thing that eventually gets us to the place that we think and hope this technology is going, but that actually works today.Have you figured out how to sort of sequence your way to this magical thing that may someday be t ...

AI在工业领域的应用 - 工业领域拥有全球 70% 的劳动力，但生成式AI在工业领域的应用却被忽视[2][3] - IFS 认为 AI 在工业领域的应用比在企业办公室更具影响力，能够推动工业进步和经济发展[2][4] - 通过 Boston Dynamics 的 Spot 机器人与 Eversource Energy 和 Anthropic 合作，对 5 公里的地下管道进行检测，AI 能够快速发现问题并触发工作指令，提高效率和准确性[7][8][9] - LLM 需要经过专门的训练和调整才能在特定工业环境中发挥作用，结合 IFS 的工作流和数据，可以显著提高 ROI[13][14] - IFS 扮演着连接模型、基础设施、数据、机器人和合作伙伴的角色，为工业领域的运营变革提供支持[15][16][18] - LLM 为技术人员引入了全新的互动方式，他们可以使用语音、视频和摄影等方式与平台上的智能信息进行互动，从而更好地完成工作[24] Anthropic 的合作 - IFS 选择 Anthropic 作为合作伙伴，因为 Anthropic 专注于企业应用，并重视风险与可能性的平衡，这对于需要高可靠性的工业领域至关重要[27][28][29] - IFS 对 Anthropic 抱有高度信任，但同时会进行严格的压力测试，并保持人机协作，以确保决策的准确性和安全性[30][32][33] AI 对劳动力市场的影响 - 预计 AI 将推动经济增长，未来几年全球 GDP 增长将提高 1 个百分点[42] - 世界经济论坛预测，到 2030 年，AI 将创造 1.7 亿个新工作岗位，抵消 9200 万个现有岗位的流失，净增 7800 万个工作岗位[42] - AI 和机器人技术是解决劳动力短缺问题的有效途径，特别是在老龄化劳动力和西方经济再工业化的背景下[44][46] 能源消耗与优化 - 解决能源危机的方法包括增加发电量和提高现有基础设施的效率[55] - 美国每年因电力中断损失 1500 亿美元的产出，通过减少停机时间，可以提高现有基础设施的吞吐量[55][56]

I had Mark Beni off. I had Vlad Tenev from Robin Hood on the show. They said 50% of their code is written with AI.What will that be in two years. >> 95% or 100% for sure. You'll be prompting agents.I think for us, B 44 was built in the age of LLM. And so I structured the repository in a way that's going to be very easy for LLMs to navigate through and to read the code. I think for us it's already closer to 90%. ...

RT Nick Szabo (@NickSzabo4)@NatCon2022 All the alarmist warnings lawyers make sellers of machines put in their manuals, they are now making the LLM people put in their advice. You will get many screens worth of such warnings before you get to the actual answer. ...

RT BrianEMcGrath (@BrianEMcGrath)My favorite day of the month is here again with another amazing episode of The Exponentialist with @DMattin and @RaoulGMI! They talk longevity, who is winning the LLM war and the optimization of capital for intelligence.@RealVision is running a black Friday sale, so you have no excuse to come join our community. ...

RT Avi Chawla (@_avichawla)A popular LLM interview question:"Explain the 4 stages of training LLMs from scratch."(step-by-step explanation below) https://t.co/43WiCQuJfc ...