Where Generative AI Fits in the AI and ML Landscape — Transcript

Full transcript

1. 0:06Hi, and welcome to Introduction to
2. 0:08Generative AI.
3. 0:10Don't know what that is? Then you're in
4. 0:11the perfect place.
5. 0:13I'm Roger Martinez, and I am a developer
6. 0:15relations engineer at Google Cloud. And
7. 0:17it's my job to help developers learn to
8. 0:19use Google Cloud.
9. 0:21In this course, I'll teach you four
10. 0:23things.
11. 0:24How to define Generative AI.
12. 0:27Explain how Generative AI works.
13. 0:29Describe Generative AI model types.
14. 0:31Describe Generative AI applications.
15. 0:34But let's not get swept away with all of
16. 0:36that yet.
17. 0:37Let's start by defining what Generative
18. 0:40AI is first.
19. 0:42Generative AI has become a buzzword. But
20. 0:44what is it?
21. 0:45Generative AI is a type of artificial
22. 0:48intelligence technology that can produce
23. 0:50various types of content, including
24. 0:52text, imagery, audio, and synthetic
25. 0:56data.
26. 0:57But what is artificial intelligence?
27. 1:01Since we are going to explore Generative
28. 1:03Artificial Intelligence, let's provide a
29. 1:05bit of context.
30. 1:07Two very common questions asked are What
31. 1:11is artificial intelligence? And what is
32. 1:13the difference between AI and machine
33. 1:15learning?
34. 1:17Let's get into it. So one way to think
35. 1:19about it is that AI is a discipline,
36. 1:21like how physics is a discipline of
37. 1:23science.
38. 1:24AI is a branch of computer science that
39. 1:26deals with the creation of intelligent
40. 1:28agents and our systems that can reason,
41. 1:31learn, and act autonomously.
42. 1:34Are you with me so far?
43. 1:35Essentially, AI has to do with the
44. 1:37theory and methods to build machines
45. 1:39that think and act like humans.
46. 1:42Pretty simple, right?
47. 1:44Now, let's talk about machine learning.
48. 1:47Machine learning is a subfield of AI.
49. 1:50It is a program or system that trains a
50. 1:52model from input data.
51. 1:55The trained model can make predictions
52. 1:57from new, never-before-seen data drawn
53. 2:00from the same one used to train the
54. 2:01model.
55. 2:03This means that machine learning gives a
56. 2:04computer the ability to learn without
57. 2:07explicit programming.