正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

Wow! We just put a human-in-the-loop within ChatGPT. ❤️🤖

Marc Benioff

3,112,223 subscribers

104,962 次观看 • 7 个月前 •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 条评论

暂无评论

原始帖子的评论将显示在这里

相关视频

2025. The year humans trained robots. 🤖 70,000+ teleop sessions 🤖 Nearly 5,000 hours of real human-in-the-loop work 🤖 1.2M+ app members This is PrismaX Wrapped 2025.

2025. The year humans trained robots. 🤖 70,000+ teleop sessions 🤖 Nearly 5,000 hours of real human-in-the-loop work 🤖 1.2M+ app members This is PrismaX Wrapped 2025.

PrismaX

54,473 次观看 • 5 个月前

🤖Simulation-Ready Human–Scene Interaction Reconstruction🤖 #HSImul3R reconstructs sim-ready 3D human-scene interactions from casual videos with *physics-in-the-loop* that can be directly deployed to humanoids. - Project: - Code:

🤖Simulation-Ready Human–Scene Interaction Reconstruction🤖 #HSImul3R reconstructs sim-ready 3D human-scene interactions from casual videos with physics-in-the-loop that can be directly deployed to humanoids. - Project: - Code:

Ziwei Liu

15,326 次观看 • 3 个月前

What we do with "ChatGPT" 🤖

What we do with "ChatGPT" 🤖

Felix

14,998 次观看 • 4 个月前

Sometimes the model’s driving. Sometimes the human is. With Lite-T, a wearable data collection system, operators can jump in and out of control during training—seamless human-in-the-loop. Tag team training for embodied AI 🤝🤖 #data #embodiedai

Sometimes the model’s driving. Sometimes the human is. With Lite-T, a wearable data collection system, operators can jump in and out of control during training—seamless human-in-the-loop. Tag team training for embodied AI 🤝🤖 #data #embodiedai

Galaxea Dynamics

21,113 次观看 • 3 个月前

New: you can now use your ChatGPT subscription within Roo Code’s loop, officially supported by OpenAI. We're open source so this is the kind of user choice we ❤️ use what you already have, inside your existing workflow. Here's how to set it up in your IDE

New: you can now use your ChatGPT subscription within Roo Code’s loop, officially supported by OpenAI. We're open source so this is the kind of user choice we ❤️ use what you already have, inside your existing workflow. Here's how to set it up in your IDE

Roo Code

39,957 次观看 • 5 个月前

Put this on a loop and just vibe.

Put this on a loop and just vibe.

Forza Horizon

107,093 次观看 • 2 年前

ChatGPT learned from human feedback. The next generation of AI will learn from the laws of nature. Liam Fedus (co-creator of ChatGPT, now at Periodic Labs) says models will train through “experiment in the loop”, using real-world results, not human preference, as their reward function. William Fedus Periodic Labs

ChatGPT learned from human feedback. The next generation of AI will learn from the laws of nature. Liam Fedus (co-creator of ChatGPT, now at Periodic Labs) says models will train through “experiment in the loop”, using real-world results, not human preference, as their reward function. William Fedus Periodic Labs

a16z

72,646 次观看 • 8 个月前

I rode in the new self driving Tesla in the Las Vegas Loop! It was WAY smoother than a human driving in the Loop. Zero human input. It wasn’t busy so he let us ride back as well. Driver said they soon will be giving rides all the way to the airport. We are in the future.

I rode in the new self driving Tesla in the Las Vegas Loop! It was WAY smoother than a human driving in the Loop. Zero human input. It wasn’t busy so he let us ride back as well. Driver said they soon will be giving rides all the way to the airport. We are in the future.

TesAli

296,844 次观看 • 7 个月前

“The energy we create in a room together is just… wow” ∞

“The energy we create in a room together is just… wow” ∞

juli ☾

23,744 次观看 • 1 年前

The MOMENT I’m put in an infinite loop on a fighting game…

The MOMENT I’m put in an infinite loop on a fighting game…

City

149,929 次观看 • 1 个月前

"We can put down any world leader, anywhere in the world, anytime we want within a five-day window if we wanted to."

"We can put down any world leader, anywhere in the world, anytime we want within a five-day window if we wanted to."

Carl Higbie

90,433 次观看 • 5 个月前

The best human robots show ever - by a wide margin. The Chinese new year TV this week put the human robot technology to full display. Wow - the robots can dance, box, fect, do gymnastics, turn around, interact with children - please watch! It’s simply amazing - and we all realize - this is just the start.

The best human robots show ever - by a wide margin. The Chinese new year TV this week put the human robot technology to full display. Wow - the robots can dance, box, fect, do gymnastics, turn around, interact with children - please watch! It’s simply amazing - and we all realize - this is just the start.

Erik Solheim

75,926 次观看 • 4 个月前

AI shouldn’t be used to spy on Americans. AI shouldn’t be used in deadly autonomous weapons without a human in the loop. This is just common sense.

AI shouldn’t be used to spy on Americans. AI shouldn’t be used in deadly autonomous weapons without a human in the loop. This is just common sense.

Sen. Elissa Slotkin

15,579 次观看 • 3 个月前

Wow. Adam Mockler just put on a masterclass. Watch till the end.

Wow. Adam Mockler just put on a masterclass. Watch till the end.

MeidasTouch

220,572 次观看 • 8 个月前

🤖ALPHABOT SOCIETY TURNS 1🤖 In honor of hitting such a big milestone, we put together a recap highlighting some of what building this project has been like over the last year for your viewing pleasure🙏 This has been an incredible year and we are just getting started! #CROFam

🤖ALPHABOT SOCIETY TURNS 1🤖 In honor of hitting such a big milestone, we put together a recap highlighting some of what building this project has been like over the last year for your viewing pleasure🙏 This has been an incredible year and we are just getting started! #CROFam

AlphaBot Society🤖

22,088 次观看 • 3 年前

ChatGPT just launched a new feature!! You can @-mention any custom GPT from within ChatGPT—almost like Slack or Discord. It's a gamechanger! Walkthrough here:

ChatGPT just launched a new feature!! You can @-mention any custom GPT from within ChatGPT—almost like Slack or Discord. It's a gamechanger! Walkthrough here:

Dan Shipper 📧

286,489 次观看 • 2 年前

Within the span of 72 hours, we have had two tropical cyclones and their remnants drop nearly a foot of rain in just a matter of a few hours. This is a 30-hour radar loop of Tropical Storm Chantal’s landfall into the Carolinas.

Within the span of 72 hours, we have had two tropical cyclones and their remnants drop nearly a foot of rain in just a matter of a few hours. This is a 30-hour radar loop of Tropical Storm Chantal’s landfall into the Carolinas.

Backpirch Weather

18,784 次观看 • 11 个月前

After digging into developer productivity, we learned something surprising: “human in the loop” is mostly just waiting. 💤 🔥 So we fixed it. IntelliJ IDEA is now a gaming platform 🎮️ Start playing 👉️ P.S. Codex is working on the next version as we speak.

After digging into developer productivity, we learned something surprising: “human in the loop” is mostly just waiting. 💤 🔥 So we fixed it. IntelliJ IDEA is now a gaming platform 🎮️ Start playing 👉️ P.S. Codex is working on the next version as we speak.

IntelliJ IDEA, a JetBrains IDE

22,826 次观看 • 2 个月前

#GetAgent just got even smarter! 🚀 🤖 Access anytime with a floating window 🤖 Personalized saving & trading bot product recommendations 🤖 Closed-loop futures trading (coming soon) Try it now and level up your trading with the #BestCryptoAI!

#GetAgent just got even smarter! 🚀 🤖 Access anytime with a floating window 🤖 Personalized saving & trading bot product recommendations 🤖 Closed-loop futures trading (coming soon) Try it now and level up your trading with the #BestCryptoAI!

Bitget

16,751 次观看 • 8 个月前

If we wait for 100% safety before OK'ing new technologies (e.g. AI therapists) we end up withholding enormous benefits from the very people they’re meant to help. The benchmark is not zero mistakes; it's a system that is safer than just having a human in the loop.

If we wait for 100% safety before OK'ing new technologies (e.g. AI therapists) we end up withholding enormous benefits from the very people they’re meant to help. The benchmark is not zero mistakes; it's a system that is safer than just having a human in the loop.

Reid Hoffman

37,542 次观看 • 7 个月前