Jason R. Dyck

Changing the world, from the inside out.

[Andrej Karpathy @karpathy]

This is the the quote I've been citing a lot recently.

[Jason R. Dyck @IslandJayD]

@koltregaskes API is my guess.

[Jason R. Dyck @IslandJayD]

@SullyOmarr Hard to believe that the average person doesn't even notice, and is still unsure about AI. 🤯

Meanwhile, the world is shifting in large ways daily.

[Jason R. Dyck @IslandJayD]

@itsandrewgao Right? How about if it saves an hour of time working with a lawyer?

I've been using pro avidly to create and fine tune a data set for a an app "Kids Need Dads".

The education in creating this tuned data set alone is worth the $200. And it helped with the application on day 2!

[Jason R. Dyck @IslandJayD]

@deedydas Makes me wonder how that Swarm trend will go, being that it's so early in its lifecycle.

[Jason R. Dyck @IslandJayD]

@stevenbjohnson You bet! Interacting with my 2 year journal was one of the first things I used it for.

What a great experience in being witnessed and honoured through the deep dive reflection on it.

[Jason R. Dyck @IslandJayD]

Phenomenal addition, now supporting audio files too!

[Steven Johnson @stevenbjohnson]

How to do school with NotebookLM:

1. Record audio from class on your phone
2. Keep laptop closed. Just jot down short phrases to describe most important points
3. Upload audio and PDF scan of notes to NotebookLM
4. Ask Notebook to expand your notes with details from recording

Bonus: at the end of the week, create an Audio Overview from all your class summaries to review the most important concepts in podcast format.

(As of this morning, NotebookLM now supports audio files--and YouTube videos--as sources. And we've added easy sharing tools for Audio Overviews.)

https://t.co/6hhNSzBKOZ

[Jason R. Dyck @IslandJayD]

@UltraRareAF Right? Once we straighten out some of the current guardrails at least. I trust its capability more than many in leadership roles.

[Jason R. Dyck @IslandJayD]

Great use case, these tools can augment learning immensely.

[Steven Johnson @stevenbjohnson]

How to do school with NotebookLM:

1. Record audio from class on your phone
2. Keep laptop closed. Just jot down short phrases to describe most important points
3. Upload audio and PDF scan of notes to NotebookLM
4. Ask Notebook to expand your notes with details from recording

Bonus: at the end of the week, create an Audio Overview from all your class summaries to review the most important concepts in podcast format.

(As of this morning, NotebookLM now supports audio files--and YouTube videos--as sources. And we've added easy sharing tools for Audio Overviews.)

https://t.co/6hhNSzBKOZ

[Mckay Wrigley @mckaywrigley]

My #1 workflow change recently is adding comments *everywhere* in my code with Cursor.

Want a major codegen performance boost?

Then document your codebase to drastically help the AI that’s working on it.

It’s super easy - prompt below.

[Jason R. Dyck @IslandJayD]

@karpathy This has me wonder at the future of conveying our thoughts and ambitions in a manner other than language.

Whether Neuralink or something else entire, it seems like going beyond words could significantly increase the value we create with LLMS.

Thoughts?

[Jason R. Dyck @IslandJayD]

Try putting these or any white papers into https://t.co/iVsXOCJQPp (short waitlist) for an audio discussion of that core concepts.

Making dense white papers far more accessible.

Or try it in https://t.co/2LHpOKl2Us if you want to have a casual voice to it

#AI #whitepapers

[Philipp Schmid @_philschmid]

5 papers you want to read to understand better how @OpenAI o1 might work. Focusing on Improving LLM reasoning capabilities for complex tasks via training/RLHF, not prompting. 👀

> Quiet-STaR: Language Models Can Teach Themselves to Think Before Speaking (https://t.co/ote01ei0hQ) from Stanford

> Agent Q: Advanced Reasoning and Learning for Autonomous AI Agents (https://t.co/ZkGtrzRcV9) from MultiOn/Stanford

> Let's Verify Step by Step (https://t.co/g2iPYVMd6N) from OpenAI

> V-STaR: Training Verifiers for Self-Taught Reasoners (https://t.co/34VLwce0HZ**) from Microsoft, Mila**

> Learn Beyond The Answer: Training Language Models with Reflection for Mathematical Reasoning (https://t.co/BJE5u62B39) from Notre Dam, Tencent

I'm not claiming this is how O1 works, but it helps us better understand it. I'll share summary posts in the coming days. Make sure to follow! 🫡

[Jason R. Dyck @IslandJayD]

@elder_plinius Whoa!!!

[Jason R. Dyck @IslandJayD]

@wesroth Sounds like it, doesn't it?

[Jason R. Dyck @IslandJayD]

@lexfridman "If you lead the people with regulations and keep them in order with punishments, they will avoid punishment but will have no sense of shame. If you lead them with virtue and keep them in order with rites, they will develop a sense of shame and moreover will order themselves."

[Jason R. Dyck @IslandJayD]

@lexfridman "Good people do not need laws to tell them to act responsibly, while bad people will find a way around the laws." - Plato

[Jason R. Dyck @IslandJayD]

@lexfridman "The more laws and restrictions there are, the poorer people become." - Tao te Ching

This wisdom shows up in many places.

[Jason R. Dyck @IslandJayD]

@UltraRareAF I had fun dropping a link to this tweet into https://t.co/2LHpOKl2Us and listening to a deep dive on it.

Super fun have AI hosts discuss AI sentience. :D

[Jason R. Dyck @IslandJayD]

Watching this closely and hoping it turns out to be as impactful as I anticipate it might me.

[Jason R. Dyck @IslandJayD]

@mckaywrigley Big time!

All of these systems learning form one another is going to ramp quickly and to great extents.