Victor.rsl

nothing

[Victor.rsl @Victorrsl1]

@LLLucifEEErR 爱情的罪证😆

[Victor.rsl @Victorrsl1]

@GuoShenCN7 老板你好，我申请回复员/管理员职位。

[小互 @xiaohu]

2024 Inclusion·外滩大会在上海开幕，著名未来学家凯文·凯利在开幕主论坛上发表演讲：当人工智能深刻影响经济和文化时，将涌现三大趋势！

非常值得观看，强烈推荐！

AI时代三大趋势：

• 全球主义：

• 全球正在共同构建一个基于技术的“超级有机体”，将全球的设备连接成巨大的计算系统，所有数字技术包括人工智能都运行在这个新平台上。

• 全球主义推进推动了新兴的全球文化，人们的生活方式和穿着逐渐趋同，底层需求达成一致。

• 人工智能将实现真正的“实时翻译”，打破全球语言障碍，加上增强现实（VR）技术，全球化“劳动力”将第一次出现。

• 创新加速：

• AI技术加速了创新的步伐，体现在新发明和新思想传播速度加快，通过增强现实（AR）和虚拟现实（VR）技术提高学习效率，AI通过机器及其他传感器感知世界，以及ChatGPT等人工智能工具加快人们学习速度。

• 当答案易得时，提出正确问题和掌握正确思维方式变得更有价值，“学会如何学习”将成为未来毕业生的核心技能。

• AI驱动生成：

• 人工智能能帮助人们摆脱机械、枯燥的工作，完成无法独立完成的任务，甚至生成超乎人类想象的全新事物。

• 人工智能系统正在生成新事物且不断完善，人工智能是多样的，将填补许多智能领域。

• 从人工智能工具中获益最多的是表现一般或较差的员工，人工智能工具能帮助他们更好地完成工作。

• 回答中国网友提问：

• 对于人工智能是不是人类的下一个进化阶段，凯文·凯利给出肯定答案，认为其重要性不亚于火、印刷术和工业革命，但实现需要至少一个世纪甚至更长时间，人类正在通过人工智能重塑文明，创造人造“外星人”并引入新思想和文化。

• 为确保人工智能带来有益未来且不加剧不平等，应利用技术引导其发展，推动其开源化，形成更具包容性的技术生态系统。

• 未来几乎所有工作都会使用人工智能，但不会被完全取代。

[AI Will @FinanceYF5]

麻省理工大学刚刚发布了免费的在线课程。

无需申请或支付费用。

以下是你在2024年不想错过的10门课程：

[Ryohei Sasaki@engineer @rsasaki0109]

control_barrier_function
Control Barrier Function Example
https://t.co/FsNm1w31h7

[Victor.rsl @Victorrsl1]

thanks for sharing

[Yohan @yohaniddawela]

Traffic data is one of the best predictors of local economic activity.

But it's notoriously difficult and expensive to access.

There's now a free traffic dataset which covers 40 cities.

Here's what you need to know about it:

[Takahiro Yabe @takayabe_0505]

📢 Dear colleagues working on human mobility modeling and prediction -- we are excited to announce a new open dataset of human mobility which contains city-scale, longitudinal, anonymized, 100,000 individual mobility trajectories. Scientific Data: https://t.co/IsgtviS9SC

[Tony S.F. @tonysilveti]

Made this when I was bored and I don't care to refine it more so here you go - visualizing how the frank-wolfe algorithm works on a simple example: the constraint set is an octagon in R^2 and the objective function is just the squared distance to the point a (denoted by a star).

[Conor Heins @conorheins]

Our paper "Collective behavior from surprise minimization" now published in PNAS @PNASNews! Check it out to learn about a new, flexible approach to modelling collective motion with active inference!

[Active Inference Institute @InferenceActive]

Here is a 714 GB torrent that contains a snapshot of the public videos produced by the Active Inference Institute on through March 2024. https://t.co/bJZPSKzo5v
We'll be looking forward to improve the quality of these distributions (in terms of more nuanced and documented torrents). For now for this first one, it is just a bulk download.
Enjoy & give us feedback as you see relevant ~

[Conor Heins @conorheins]

New preprint! “Collective behavior from surprise minimization,” by me, @BerenMillidge, @lancelotdacosta, @richardpmann, Karl Friston, @icouzin
Paper: https://t.co/x6Qv2XtaUc
Code: https://t.co/DfK8xSgzHH
🧵 (1/14)

[Giorgio Nicoletti @gnicoletti09]

Our latest work with @dbusiello1 is out in Physical Review X! We combine ideas from multilayer networks, higher-order interactions, and information theory to study how stochastic processes are coupled across timescales @PhysRevX

A long 🧵

https://t.co/u6jPP1zTuL

[Yen-Huan Li @yenhuan_li]

==== My recommendations today ====

Faster Convergence of Stochastic Accelerated Gradient Descent under Interpolation
https://t.co/rzI3lDEo0O

Sum-of-Squares Relaxations for Information Theory and Variational Inference
https://t.co/yLOzfJ6Gbe

(1/n)

[Tsarathustra @tsarnick]

Sam Altman says that content curation and the ability to figure out what people want will be the most valuable skills in an AI-enhanced future

[Stefano V. Albrecht @s_albrecht]

Excited to announce that China Machine Press will publish the Chinese translation of our new MARL textbook in Dec 2025 📚 Looking forward to more audiences being able to reach the fascinating world of Multi-Agent Reinforcement Learning 🎉 @mitpress https://t.co/Q4KaoywQQz

[Texas ASE/EM @UTAerospace]

Tired of sitting in rush hour traffic?🚦🚗David Fridovich-Keil’s research to develop #GameTheory models for improved #traffic flow just might be the answer. The project is being funded by a new @NSF CAREER award. Learn more: https://t.co/ZzqDbDmMHm

[OpenAI @OpenAI]

We’ve built an energy-based model that can quickly recognize, generate, and transfer simple concepts after only 5 training demos: https://t.co/JD5Lt8ZYIb

[Gabriel Peyré @gabrielpeyre]

Gradient descent is consistent with a first-order ODE. Nesterov’s acceleration is consistent with a second-order ODE. Nesterov is faster in the worst case but gradient is better for strongly convex functions. https://t.co/nysptAoktl

[Machine Learning Street Talk @MLStreetTalk]

Jeff Beck comparing active inference to reinforcement learning, and the steerability of multi agent systems.

[Victor.rsl @Victorrsl1]

Check out this blog post from @LouisKirschAI https://t.co/DIdZlYHLSD