Machine Learning vs. Deep Learning vs. Foundation Models

Where does NLP fit in?

Superb explanation

Thank you for bringing this video

Here is what I have noted till so far.. AI>= ML>=DL
Machine learning
—Supervised and Unsupervised
Supervised—Human intervention ,no need of labeled datasets
Unsupervised—no human intervention, machine labels data itself
Deep Learning
—Neural networks, more than 3layers, and learns itself


Supervised—-trained on labeled data… same like saying child anything with 4 legs(4 lines) is cat, now we need more label for accuracy, anything with meow sound…
Unsupervised—machine learn itself by finding patterns in data and sub classifying them,, it is like child without aid of parent start recognizing things itself…parents can aid in correcting mistakes in between/ With Feedback from environment -Reinforced Learning


Deep Learning—- Specifically focuses on artificial neural networks with multiple neural layers

Machine Learning—also includes other techniques like linear regression,decision tress,support vector machine, clustering algorithms



How i am able to speak in proper grammar structure ,that is also cohesive
Because after speaking for so many times with lot of mistakes and improvisation, i have developed a pattern (grammatical and cohesive)pattern that will decides COHESION of my speaking..
I have learnt a lot of words,meanings from environment;(books,audio,people ) and figured out the meaning and when to use it based on the CONTEXT..
Right words in right context with right grammatical cohesion


Foundational Models
Large scale neural network models, already trained on vast amount of diverse datasets for a particular task or purpose(like language,audio recognition and generation, content generation)
So instead of training model from scratch,This pre trained task specific models act as foundation for multitude of applications… like
- [ ] Language Translation
- [ ] Content Generation
- [ ] Audio Recognition
- [ ] Image Recognition
- [ ] Large Language Model—— processing the language( understanding the grammar, context, idioms, sentiment of the language through human alike TEXT) and generating replies, reasoning, translation, paraphrasing,
- [ ] Vision Models—- Recognize image, classify and interpret it.. For example when we see a object, first we classify it what it is, what it is doing with its body features( for example arms spread— human fighting) so not only recognizing the image through facial, color and body structure pattern but also interpreting the context within that image.. Now when we can recognize cat, we can draw, generate cat, when we know how cat jumps via image, we can draw cat jumping…RECOGNIZE,INTERPRET AND GENERATE IMAGES
- [ ] Scientific models— biology
Now Foundation Model helps in understanding and interpret information , Generative AI by inferring the vast knowledge from this foundation models, creatively generates language text, audio, image based on the prompt given

Thank you, great explanation. One remaining question - where does “Data Science” fit? Do you see it as encapsulating all the boxes, plus a little more? And if so, what is the “more”?

Where is NLP located?

Beautifully explained. Thank you.

where does hugging face and cohere fall?

Enormity isn't size, it's more like being horrorified.

How do you write so well backwards on the glass?

Not sure I agree that RL belongs under ML

Hello, what about data science

There's a huge circle that encapsulates all the boxes and it's called tooling. Not sarcastic.

What is there under AI, other than Machine Learning?

Learnt a new term claro. Like that and this. Great explanation!

How valuable is data, authentication for the training of these tools, refined thoughts, at rapid speed.

Would a new supply chain movement towards generating a new standardize benchmark system, be useful? Potential sufficient to correct the potential errors, of miscommunication via scholarly debate. Perhaps chaos, but perhaps the cure. 😅 all in the amount of effort

eXcellent. Thank you.

Great way to start the day
💪🤖

I love this guy's energy, very informative