筛选结果
22 篇排序
10:36EN → 简体中文Meta Muse Wonder 实测：基准成绩强劲，演示表现不一，额度限制影响体验Matthew Berman · Forward FutureMatthew Berman 测试 Meta 模型的浏览器演示，并分析公开的基准成绩、使用方式和成本。
0:24EN → 简体中文Claude Fable 5 访问期限延长至 2026 年 7 月 12 日Matthew Berman · Forward FutureMatthew Berman 报道 Claude 付费订阅者获得短期延期，但原有使用额度不会重置。
13:30EN → 简体中文
3Blue1Brown· 2025年8月17日AlphaGeometry 如何结合逻辑证明与创造性作图Aleph0Aleph0 讲解了符号几何求解器如何在不使用 AI 的情况下解决大量难题，以及语言模型如何提供辅助作图，让系统达到奥数竞赛水平。
3:13EN → 简体中文Claude Opus 4.1：Anthropic 在 2025 年 8 月的基准更新Matthew Berman · Forward FutureMatthew Berman 回顾模型报告的编程、推理和工具使用结果，其中一项基准出现下降。
12:14EN → 简体中文AI 新闻：Windsurf 收购、Meta 算力计划、Grok 与 VoxtralMatthew Berman · Forward Future一份 2025 年 7 月的新闻汇总，讨论开发者工具交易、AI 基础设施、模型行为、奖励方法和语音识别。
1:09EN → 简体中文关于 GPT-4.5 规模与训练的讨论一位发言者评价 GPT-4.5 的成本、速度与据称遇到的训练挑战。
1:50EN → 简体中文Coveo 的 Sébastien Paquet 讨论的五个 AI 趋势Matthew Berman · Forward FutureMatthew Berman 总结一场 2025 年访谈，涉及生产系统、小模型、智能体与信任。
0:35EN → 简体中文探访 IBM：Granite 模型、InstructLab 与量子计算Matthew Berman · Forward FutureMatthew Berman 预告一段关于 IBM 小型 AI 模型、模型定制和量子研究的长视频。
0:54EN → 简体中文初探 OpenAI o1 Preview：让模型花更多时间推理Matthew Berman · Forward FutureMatthew Berman 介绍 o1-preview 和 o1-mini，并尝试用 Python 游戏提示词测试模型。
9:43EN → 简体中文Matthew Berman 测试 Jamba 1.5 的编程与推理能力Matthew Berman · Forward Future在 2024 年的一次上手测试中，AI21 的混合架构模型输出缓慢，在少量提示词上的表现有好有坏。
0:58EN → 简体中文NVIDIA NIM：容器化 AI 推理微服务Matthew Berman · Forward Future一段赞助内容介绍 NVIDIA 如何将模型推理打包，供本地和云端应用使用。
12:50EN → 简体中文测试 Smaug：托管 70B 与较小本地模型的比较Matthew Berman · Forward FutureMatthew Berman 在贪吃蛇、数学和推理题上比较两种 Smaug 配置。
14:23EN → 简体中文实测 DBRX：编程能力亮眼，推理仍有短板Matthew Berman · Forward FutureMatthew Berman 介绍 Databricks 的开放权重模型，并用编程、算术和推理题进行测试。
5:24EN → 简体中文Google I/O 2023：Bard、PaLM 2 与 Google 产品中的 AIMatthew Berman · Forward FutureMatthew Berman 回顾 Google 2023 年的 AI 发布，分析其重点改进现有产品的策略。
9:59EN
Lex Fridman播客Is Language or Vision Harder? Ilya Sutskever on the Unity of Machine Learning主持 Lex FridmanIn a short clip from his 2020 conversation with Lex Fridman, Ilya Sutskever explains why vision, language and reinforcement learning share the same few principles, and why asking which problem is harder is the wrong question.
11:11ENLex Fridman播客 · 2 人对谈Yann LeCun on Benchmarks, Interactive Environments, and Why Human Intelligence Is Not General嘉宾 Yann Lecun · 主持 Lex FridmanIn a 2019 clip, LeCun defends benchmarks as a test of AI claims, explains why interactive environments break the usual train-test protocol, and argues that "general intelligence" is a misnomer.
10:16ENLex Fridman播客 · 2 人对谈Reasoning Without Logic: Yann LeCun on Teaching Neural Networks to Think嘉宾 Yann Lecun · 主持 Lex FridmanYann LeCun explains why he believes neural networks can reason, and what memory, recurrence and continuous representations would need to look like.
13:58ENLex Fridman播客Deep Reinforcement Learning, Explained by the Researcher Behind AlphaGo主持 Lex FridmanDavid Silver of DeepMind walks through what reinforcement learning is, how its building blocks fit together, and why deep neural networks keep surprising even the experts.
10:34ENLex Fridman播客 · 2 人对谈Why Deep Learning Needs Rules: Chollet on Interpolation, Abstraction and Hybrid AI嘉宾 François Chollet · 主持 Lex FridmanFrançois Chollet explains why neural networks only interpolate between points they have seen, and why the best AI systems pair them with symbolic rules.
10:39ENLex Fridman播客Statistics as Decision-Making: Michael I. Jordan on Bayesians, Frequentists and the Ideas Between Them主持 Lex FridmanBerkeley's Michael I. Jordan explains where statistics came from, why it is really about decisions, and how empirical Bayes and false discovery rate bridge two rival philosophies.
11:29ENLex Fridman播客Why Tesla AI Day Was the Most Impressive Real-World AI Effort Lex Fridman Has Seen主持 Lex FridmanLex Fridman walks through the neural network, data annotation, Dojo and Tesla Bot announcements and what they say about solving real-world robotics.
11:23ENLex Fridman播客 · 2 人对谈Science Is a Self-Improving System That Isn't Exploding: Chollet on the Limits of an Intelligence Explosion嘉宾 François Chollet · 主持 Lex FridmanFrançois Chollet uses the history of science to argue that recursive self-improvement meets exponential friction and yields linear, not explosive, progress.