Notes on The Turing Lectures: "What is generative AI?"

 The Turing Lectures: What is generative AI?

Generative - create new content (audio, code, images, text, video)

Artificial Intelligence - automatically using a computer program "a lot of myth about it, its just a tool", "we have to spell it out"

Generative AI is not a new concept. Google translate (17 yo) is an example of generative AI.

Another example is Apples Siri (11yo), another generative AI. Phone autocomplete or Google search autocomplete

Not that new so "what is the fuss"? Answer: chatGpt was launched. chatGpt is a lot more sophisticated than older AIs. You can have a conversation.

The technology is not extremely new. It uses language modeling. Given a context, the language model can predict

e.g. context: "I want to"

  • prediction: play, eat, shovel


AIs used to count to predict but latest use Neural Networks. Feed context (prompt text) into neural language model which predicts the answer.

Basic steps to build a language model:

  1. Step 1. Collect very large data (aka "corpus")  e.g. the web (wikipedia books, github, stack overflow)
  2. Step 2. Ask language model to predict the next word in a sentence
    • randomly truncate last part of input sentence, calculate probability of missing works, adjust and feedback to the model
    • e.g. "the trevi fountain is in ??"  Rome - good  Berlin - bad
  3. Step 3. Repeat over whole corpus. Keep going over months.
    • So it learns over time by practicing and correcting.


Neural Network Language Models

Parameters = # input units * other nodes


Transformers - "king of AI architectures"

 - uses mini neural networks


gpt: "generative pretrained transformer"

pre-training how is it done?

  • hire expert trainers who train the AI (correct answers)
  • in the wild, how people respond and provide multiple replies for users to choose


"Fine tuning" - specializing the network e.g. add medical data to generic pretrained model

gpt is fine tuned to be general purpose


How can a model become great? "Size matters, bigger is better"

Since 2018 model sizes have grown. gpt1 "ant brain" ...now gpt4 has 1 trillion parameters  "more than rat brain"

The more parameters, the more powerful the AI. When you get to half a billion parameters then "common sense reasoning" and "summarization" and "arithmetic" etc become possible and powerful

gpt4 has read almost all the human written text

gpt4 cost $100 million, openAI can do it because they have Microsoft backing them (not everyone can)


"HHH" framework i.e. Helpful, Honest and Harmless

  • How do you make the model HHH? You Fine tune it by letting it know if answers are correct or not.


LLMs generate hallucinations by fabricating nonexistent or false facts


Llama 2, a chatGpt like model from Meta

Comments

Post a Comment

Popular posts from this blog

My Reading Lists

I had some time over Christmas and decided to organize my reading lists. Interesting how some time away from daily work allows one to raise up your head   Uncle Bob on Principles of OOd See also book: Agile Principles, Patterns, and Practices in C# Available on Safari Books online. Algorithms in a Nutshell Nine   Algorithms   That Changed   the Future Programming Pearls book: http://www.cs.bell-labs.com/cm/cs/pearls/toc.html 97 Things Every Programmer Should Know Seven L a nguages http://pragprog.com/book/btlang/seven-languages-in-seven-weeks Codys book : http://www.domenlightenment.com/ Principles of OO programming: http://www.nczonline.net/blog/2012/12/18/now-available-principles-of-object-oriented-programming-in-javascript-beta/ Object Oriented programming with Stefan Stoy... Chrome V8 : http://code.google.com/p/v8/ https://developers.google.com/v8/ https://developers.google.com/v8/design - read this Mozilla Javascript page ...
Read more

angular js protractor e2e cheatsheet

Many from the protractor test suite category detailed task syntax example load page load a page browser.get('yoururl') browser btn actions browser back/forward browser.navigate().back() OR browser.navigate().forward() access elements on page See Protractor API list and examples from Protractors test suite access by angular ngmodel name (must match ngmodel in the html) this is the preferred approach element(by.model('person.name')); access by the css id on the page e.g, #user_name (but this can be brittle) element(by.id('user_name')) access angular binding to get generated value i.e. for {{}} element(by.binding('person.concatName')); access angular repeat list data (must match repeat stmt in the html) See Protractor test suit l...
Read more

react-select stacking order bug, z-index, layers and stacking

Image
Had to fix a tricky bug related to react-select recently (our Dropdown component wraps react-select). Here's the how the bug manifested. Dropdown does not overlay dynamically added content below it Variations of this have been reported since 2017. But I tried all the suggestions in the comments and none worked for me. I also tried wrapping the Dropdown in a div and changing z-index to be higher as well as making absolute position but no success. Changing the elements order in the html source and then reordering using flexbox order does work but it breaks the tabbing flow and is not recommended from an accessibility standpoint. One issue is the Dropdown is in a container and is constrained by the containers z-index. No matter how I change the Dropdowns z-index, its container is stacked below the next in stacking order. Google Chrome debugger layers tool provides a cool way to visualize layers. Could be better but helps, check it out below. You have to look closely but you can see h...
Read more