Here's a engaging summary of the TH-cam lecture you linked, focusing on the key highlights of Natural Language Understanding (NLU): **Natural Language Understanding: A Look Back and a Leap Forward 🚀** * **The Exciting Evolution of NLU:** Professor Manning kicks off the lecture by highlighting how far Natural Language Understanding has come. Starting from 2012's early wave of interest to today, the field has exploded! This makes it a super interesting time to study NLU, as things are changing fast! * **NLU's Amazing New Skills:** The things we can do with language models today are honestly mind-blowing. They can: * Write different styles of text (think poems, code, scripts... you name it!) ✍ * Translate between languages like a pro 🌎 * Answer your questions in actually helpful ways 🤔 * **The Tricky Question Test:** One way to see NLU progress is to ask the same question to models over time. The question "Which U.S. states border no U.S. states?" is surprisingly hard due to the word 'no'. Here's how models tackled it: * 1980: SHRDLU, a classic system, could reason about complex situations...but failed on basic things outside its knowledge 🙅♂ * 2009: Wolf... pretty much just listed all U.S. states 😅 * 2020-2021: OpenAI models start getting the idea of 'no other states', but make some geographical blunders (Puerto Rico isn't a state!) * 2022: DaVinci nails it! Alaska and Hawaii are the champs of isolation 🥇 * **Challenges and the Future:** This is just a taste of NLU's journey. There are still things these models struggle with, but the field is moving rapidly. It's an awesome time to be involved in shaping where NLU goes next! ✨
9:05 - chatGPT4's answer to "Could a crocodile run the steeplechase?" The steeplechase is a track and field event in which athletes run a distance race that includes barriers and water jumps. The question of whether a crocodile could run the steeplechase is an interesting one, but there are several factors to consider: Physical Abilities: Crocodiles are not built for long-distance running. Their legs are relatively short and splayed out to the side, which is not conducive to efficient running. While they can achieve surprising bursts of speed over short distances (up to 10-12 mph for short sprints), they would not be able to maintain such speeds over the distance of a steeplechase race. Barriers: The barriers in a steeplechase would pose a significant challenge for a crocodile. Their body structure is not designed for jumping or hurdling, so they would likely have to scramble over or go around each barrier, which would slow them down considerably. Water Jumps: This is the one aspect of the steeplechase where a crocodile would excel. They are excellent swimmers and would have no problem navigating the water jumps. However, the short water sections would not make up for the time lost on the barriers and running sections. Endurance: Crocodiles are ambush predators that rely on short bursts of speed to catch their prey. They are not built for endurance running. A steeplechase race, which is typically 3000 meters for men and women in the Olympics, would be far beyond a crocodile's endurance capabilities. Safety: Even if a crocodile could be coaxed into attempting a steeplechase, it would be a significant safety risk for any humans involved. Crocodiles are wild animals and can be unpredictable and dangerous. Motivation: It's unlikely that a crocodile would have any interest in running a steeplechase race. They are not naturally inclined to run long distances or jump over barriers, and there would be no natural incentive for them to do so. In conclusion, while it's an amusing concept, a crocodile would not be suited to run the steeplechase. They lack the physical adaptations, endurance, and motivation to complete such a race.
The chart at 12:30 could be very bullish or very bearish (in a 2024 lens), depending on whether you think the lines are asymptoting at human performance or not -- synthetic data/critique/self-play seem perhaps like the way past that? Thoughts, tubers? ✊
Great question! XCS224U is the professional course: online.stanford.edu/courses/xcs224u-natural-language-understanding and CS224U is the graduate course online.stanford.edu/courses/cs224u-natural-language-understanding XCS224U was created from the graduate course CS224U. Let us know if we can answer anything else!
Do you believe is of benefit for the models to have rules and tasks? or should they be guided to learn freely instead? How can they be guided for a better improvement of society and health?
This looks an such exciting course to take right now! Thank you Stanford for releasing this!
Thank you, Prof. Potts! Thank you, Stanford!
Excellent content. Thank you, Stanford!
There is a prerequisite for this course at 54:07. Look at it first.
Thank you, I was looking for this
@@nas8318 My pleasure
Only 48k views, the world is missing out! Love the lectures
Very Deep and understandable literature on NLU. Like a story i am curious to know what will happen next.
Here's a engaging summary of the TH-cam lecture you linked, focusing on the key highlights of Natural Language Understanding (NLU):
**Natural Language Understanding: A Look Back and a Leap Forward 🚀**
* **The Exciting Evolution of NLU:** Professor Manning kicks off the lecture by highlighting how far Natural Language Understanding has come. Starting from 2012's early wave of interest to today, the field has exploded! This makes it a super interesting time to study NLU, as things are changing fast!
* **NLU's Amazing New Skills:** The things we can do with language models today are honestly mind-blowing. They can:
* Write different styles of text (think poems, code, scripts... you name it!) ✍
* Translate between languages like a pro 🌎
* Answer your questions in actually helpful ways 🤔
* **The Tricky Question Test:** One way to see NLU progress is to ask the same question to models over time. The question "Which U.S. states border no U.S. states?" is surprisingly hard due to the word 'no'. Here's how models tackled it:
* 1980: SHRDLU, a classic system, could reason about complex situations...but failed on basic things outside its knowledge 🙅♂
* 2009: Wolf... pretty much just listed all U.S. states 😅
* 2020-2021: OpenAI models start getting the idea of 'no other states', but make some geographical blunders (Puerto Rico isn't a state!)
* 2022: DaVinci nails it! Alaska and Hawaii are the champs of isolation 🥇
* **Challenges and the Future:** This is just a taste of NLU's journey. There are still things these models struggle with, but the field is moving rapidly. It's an awesome time to be involved in shaping where NLU goes next! ✨
Thank you Stanford. Happy to be among the first here 😊
Thank you for covering this old, old interest.
These lectures are awesome. Thank you stanford
Awesome set of lectures . Very knoweledgable
9:05 - chatGPT4's answer to "Could a crocodile run the steeplechase?"
The steeplechase is a track and field event in which athletes run a distance race that includes barriers and water jumps. The question of whether a crocodile could run the steeplechase is an interesting one, but there are several factors to consider:
Physical Abilities: Crocodiles are not built for long-distance running. Their legs are relatively short and splayed out to the side, which is not conducive to efficient running. While they can achieve surprising bursts of speed over short distances (up to 10-12 mph for short sprints), they would not be able to maintain such speeds over the distance of a steeplechase race.
Barriers: The barriers in a steeplechase would pose a significant challenge for a crocodile. Their body structure is not designed for jumping or hurdling, so they would likely have to scramble over or go around each barrier, which would slow them down considerably.
Water Jumps: This is the one aspect of the steeplechase where a crocodile would excel. They are excellent swimmers and would have no problem navigating the water jumps. However, the short water sections would not make up for the time lost on the barriers and running sections.
Endurance: Crocodiles are ambush predators that rely on short bursts of speed to catch their prey. They are not built for endurance running. A steeplechase race, which is typically 3000 meters for men and women in the Olympics, would be far beyond a crocodile's endurance capabilities.
Safety: Even if a crocodile could be coaxed into attempting a steeplechase, it would be a significant safety risk for any humans involved. Crocodiles are wild animals and can be unpredictable and dangerous.
Motivation: It's unlikely that a crocodile would have any interest in running a steeplechase race. They are not naturally inclined to run long distances or jump over barriers, and there would be no natural incentive for them to do so.
In conclusion, while it's an amusing concept, a crocodile would not be suited to run the steeplechase. They lack the physical adaptations, endurance, and motivation to complete such a race.
Excellent Lecture. Thank you for uploading... 💗
I don't always check my subscription feed, but I'm excited that I did today.
Awesome! Thanks for your comment, enjoy the lectures!
Me too! Thanks specifically
India attendance here 🇮🇳
🤓☝🏻❤️
The chart at 12:30 could be very bullish or very bearish (in a 2024 lens), depending on whether you think the lines are asymptoting at human performance or not -- synthetic data/critique/self-play seem perhaps like the way past that? Thoughts, tubers? ✊
Thanks specifically for your tweets link
Thanks sir
tanks
❤
what is the difference between the XCS224U cs CS224U ? The problem set and the projects assigned ?
Great question! XCS224U is the professional course: online.stanford.edu/courses/xcs224u-natural-language-understanding and CS224U is the graduate course online.stanford.edu/courses/cs224u-natural-language-understanding
XCS224U was created from the graduate course CS224U. Let us know if we can answer anything else!
Do you believe is of benefit for the models to have rules and tasks? or should they be guided to learn freely instead? How can they be guided for a better improvement of society and health?
36:00
Rolando Station