AI, Machine Learning, Deep Learning and Generative AI Explained

[@Vicinfi]

The topic is very interesting, but I'd really like to highlight the speaker, what a pleasant man! I'd like to see more content with him.

[@oy3ah2025]

Me3⭐⭐⭐⭐⭐⭐

[@Abhinayanagarajan-x8p]

"Excellent explanation of AI, Machine Learning, Deep Learning, and Generative AI! This really clears up the differences and connections between them."

[@jeffcrume]

Thanks for saying so!

[@leannyounger5625]

Hi Jeff, I’m a Contact Center Solution Architect (Business not Technical). Is there free online training (with or without a certificate) available? I need to be able to articulate this information to clients across verticals (healthcare, travel and entertainment, financial and banking etc) and know what options are available definitely technology providers (the Tech SAs that make this stuff work). Any advice/suggestions for me? 🙏😊

[@SuperCassio007]

"Back when I was an undergrad, riding my dinosaur to class" 🤣🤣

[@techwithjesus8263]

❤😂😂😂

[@jeffcrume]

I think my students at the university think it might be true 🤣

[@viswakanth9002]

I need a ride

[@PrashantGawai-zp2xt]

🤣🤣

[@Miyamoto-lo5ib]

Nice information ​@@jeffcrume

[@TanikaOmarah]

as a beginner im searching for cues, i appreciate content such as this it gives us beginners hope.

[@jeffcrume]

very glad to hear you liked it!

[@user-py4im8bd7g]

Absolutely amazing! I tried this strategy first in Demo mode with a 80 percent accuracy. Thank you for breaking this down step by step. Thank you.

[@jeffcrume]

So glad you liked it!

[@sankarnatarajan8109]

concise, simple and crystal clear articulation.

[@jeffcrume]

Thank you!

[@moderncontemplative]

Jeff Crume always breaks it down! Excellent analogy between music and LLMs. They don’t just regurgitate information just as new music is not simply regurgitated notes.

[@jeffcrume]

I’m really glad to hear that you liked it!

[@arktseytlin]

Actually a lot of it is :) and lately it's been just beats and monotony. Compare Taylor Swift to Mark Knopfler for example and you'll see that we are regressing on that front

[@XuanThaoYenTram]

This content gives beginners of AI learning hope. For those looking for something industry-specific, Immersive Translate now allows you to customize your own AI expert, it also allows translations in the technology field become more accurate and professional.

[@techwithjesus8263]

I really love how you explain things. Thanks.

[@jeffcrume]

Thanks so much for saying so!

[@thickivicki89]

I never thought I would dive this deep into AI. I am so proud of myself and all those who are taking this serious. So many people tell me learning all these technologies are a waste of time and I can't understand how some people don't think AI is apart of us for good and has been for awhile now. This video was great

[@jeffcrume]

I agree. AI is here to stay and can do a lot of good for us if we are careful to use it properly

[@AshishSharma-bc4ut]

I'm happy to be an undergrad in computer science and engineering (AI&ML). ❤

[@bagbag5138]

@@AshishSharma-bc4ut😅 ll

[@thickivicki89]

@@jeffcrume Exactly 💯

[@thickivicki89]

@@AshishSharma-bc4ut 👑🏆

[@MichaelHeinz4]

An excellent, easy way to understand the differences of certain AI methodologies & terms.....well done!!

[@jeffcrume]

Thanks for watching!

[@YvonneMcGinnis]

Always enjoy the comparative analysis between the technologies.

[@haiyinghao4037]

Thank you for the clear explanation! You make it easy for non-tech people to understand the technologies.

[@jeffcrume]

Thank you for watching!

[@1thomasdave]

A useful and welcome contribution - always great to try and agree the language of the world of AI hype - especially if we're trying to deliver value in Enterprises through 'AI'. Think this webinar might miss a key part- all those really useful ML models that aren't Deep Learning (categorization, time series forecasting, anomaly detection etc - that work on governed 'structured data'.

[@jeffcrume]

True, we try to keep these videos at or around 10 mins so I just didn’t have time to get into some of those aspects

[@MBilalSheikh05]

After watching alot of videos, I reached one of the best, Thanks sir now i am clear each models of AI.

[@jeffcrume]

I’m so glad you liked it!

[@amitconnect]

Wonderful expIaination ....i wrote programs in LIPS, Fortran and Pascal ...i did my diploma in 1998 ...I wrote the entire data structure in Pascal....then I learned c and cpp in 2000 era.

[@jeffcrume]

Glad you liked it! Nice to hear that someone else experienced those technologies

[@amitconnect]

@@jeffcrume Thanks for all hard work...out of curiosity did u teach AI ..on any edtech platform like coursera or udemy?.....your teaching style is unique why can't you create small course for interested peope...sorry I know your very busy man ...just a thoughts 💭..thanks 👍

[@sivareddy1733]

I really appreciate his level of making this crystal clear with his explanation.

[@jeffcrume]

I'm so glad to hear that you liked it!

[@sivareddy1733]

@@jeffcrume Sir, looking for more on Gen AI.

[@BriceCranford]

I love the fact that it gives good indications for placing a trade. I am always willing to learn new strategies.

[@naskar23]

Thanks for the video. Very informative. Clear and concise :)

[@jeffcrume]

I'm glad you liked it!

[@varapradha-m6r]

"Fantastic breakdown of AI concepts! SmythOS is making multi-agent AI collaboration easier and more effective. Excited to see how generative AI will augment human creativity! #AITech"

[@mahiaravaarava]

I'm particularly excited about the potential of generative AI to revolutionize content creation and problem-solving.

[@jeffcrume]

I am too!

[@NagendrababuLakavathu]

Clear and easy to understand the differences between them, Thank you so much.

[@jeffcrume]

Glad you liked it!

[@jorgesanabria6484]

Jeff Krume and Maya Murad are the best

[@jeffcrume]

Very nice of you to say so!

[@user-jr6bl9ih3e]

Should note that much of the AI advancement correlates with advancements in raw computing power. One might also want to add another bullet under AI for robotics where AI can interact with the physical world.

[@jeffcrume]

Good point. So many areas that this could all go...

[@KumR]

Thanks for recap.

[@jeffcrume]

My pleasure!

[@balddaddy3902]

I’m really impressed cause he wrote the whole explanation in a mirrored way. That’s a skill.✨

[@jeffcrume]

Sorry to disappoint but if you search the channel for "how we make them" you'll learn the trick

[@mansikanungo9396]

Use inverted video😒

[@lovewinseveryday]

More on large language models connected to Cybersecurity 😊

[@jeffcrume]

Definitely! I have some videos already on AI and cyber if you want to search the channel. I'd post a link here but it would disappear, unfortunately

[@joaovitor-nn1zy]

The videos from IBM are amazing delivering a high quality content

[@jeffcrume]

Thanks so much for saying so!

[@user-xq1wz3tp5z]

Agreed . IBM still competent in chip fab, though just at lab scale. Great research repository & tech consultancy.

[@PaulMcKillop]

Really clear and logical. Thank you

[@jeffcrume]

So glad you liked it!

[@kssadilla]

Thanks for explaining this in a visual way and using analogies ❤

[@jeffcrume]

You’re very welcome!

[@davidbeaumont6844]

Thank you for sharing in a way that is easy to digest.

[@jeffcrume]

Thank you for watching!

[@InnovativeThinkingMethods]

QUESTION: Do speakers actually write backwards on glass for these pitches, or do they write normally and use the computer to overlay the text on the video?

[@jeffcrume]

Actually, it’s neither. Search the IBM Technology channel for “how we make them” and you’ll learn the trick 😊

[@abdelazizpsy]

Thanks for the explication

[@azizmuradi4596]

Explicitly well-done. I would like to see more from you sir.

[@jeffcrume]

Thanks very much. I’ve got about 80 videos out on the IBM Technology channel to date so enjoy!

[@A032798]

Can you give one example of artificial intelligence? If finding pattern among the datasets is a ML, should AI be a more simple thing? Like routing in map, would it be considered as AI?

[@jeffcrume]

Since ML is a subfield of AI, anything you do with ML is, by definition, AI.

[@TommyeWissmann]

Thank you for being such a positive force in the trading community. Your videos have helped so many people, myself included.

[@jeffcrume]

You’re very welcome

[@petersimon985]

GM Sir, love this compact video, informative for context. Thank you for putting this together❤🎉🎉

[@jeffcrume]

Thank you for watching

[@jamesallen74]

Fantastic video. I've been looking for a video like this for a long time now.

[@jeffcrume]

So glad you liked it!

[@RickOShay]

Excellent summary and explanation!

[@jeffcrume]

Glad you liked it!

[@harper626]

very good introduction to the subject. thank you.

[@jeffcrume]

Thanks for saying so!

[@catalystx4u]

One of the Best & Finest explanation 🤩👌👌👌even a 4th grade kid would relate...

[@jeffcrume]

Thanks! I have to have make it really simple so that I can understand it in the first place 😊

[@sk3ffingtonai]

Thank you for sharing; great content, analogies, and explanation.

[@sunguroglu98]

Thank you sir.

[@master6715]

Your explanation is very clear.Thank you❤

[@jeffcrume]

Glad you liked it!

[@alimuchenik9807]

Thank you so much!!! Great video. I wonder how AIs know they are AIs. They know how they learn and understand what they know so as to be able to explain something at very different levels. Kind regards from a 67 y.o. retired woman from Argentina.

[@jeffcrume]

Hola! It's a great question which I will leave to the philosophers as I'm not sure there can be a technical answer

[@tonizeloni7520]

Thank you, again a great explanation video with lot of information combined in a simple and effective way

[@jeffcrume]

Glad you liked it!

[@rohullahkarimi8497]

clear and simple thanks

[@jeffcrume]

Glad you liked it!

[@CricketGuruji15]

Please do make a Video on Computer Vision. How can one work in that field?

[@aidsadsl999]

May I ask if you have more videos about the explanation of LLMs and practical cases in marketing strategies?

[@jeffcrume]

Yes, mostly they are from other presenters on the IBM Technology channel since my focus is on cybersecurity but take a look through the channel and you will find some

[@markfitz8315]

Thanks - always good stuff from IBM.

[@jeffcrume]

Thank you!

[@protiomertbolsonont]

We are very lucky to have you who help us in every possible way: leave a unique comment on the topic, take a screenshot to support your comment...

[@jeffcrume]

Thank you for saying so

[@covikitetamecovid1938]

I missed more information about FM, SLM & SSH for which I am very interested in. When will be your next class about these 3 parts? Thank you for your explanation Diving into simple explanation helps to better interact with the engineers communities and therefore expose them unusual design concepts for the community. Thank you.

[@covikitetamecovid1938]

I hope you read the comments early therefore you can publish soon your next content.

[@jeffcrume]

Take a look at some of the other videos on the IBM Technology channel and I think you'll find it

[@master6715]

Thank you! Your explanation is very clear.

[@jeffcrume]

Glad you liked it!

[@dimaomar9427]

Very helpful, thanks.

[@aleksandramuraveva6649]

Thank you for your video. That was amazing to watch how you built an account from $1!

[@dorins3787]

Good recap

[@maikvanrossum]

For simplicity, could you say ‘DL’ == ‘Multi-layered ML’?

[@jeffcrume]

It's more than that but if you throw in neural networks, then you get a close approximation

[@maikvanrossum]

@@jeffcrume as it’s an explainer video, how would you define neural networks, also as ‘the next step’ in development? Most of the time AI is clear, and same goes for ML. Then GenAI and LLMs are also concepts that are relatively easy to grasp. But most of the time DL as a process is equated with NNs, and then that’s it…?

[@user-qc3kj8qt3t]

​@@jeffcrume Actually we want to know how to get internship in IBM and how to build projects using these technologies if possible make a video for us...

[@vusinkosi7971]

Great content. Awesome delivery

[@jeffcrume]

Thanks!

[@proflead]

Great explanation! Thanks!

[@Staci-q8e]

Thank you. I really appreciate your good work. Your strategy using the indicator is very powerfull. Now make money is easier than ever..

[@user-uj1mf5bf7o]

perfectly explained💯

[@davidfriedman9647]

Excellent summary

[@jeffcrume]

thank you!

[@ashishshroff788]

Thank you for such great explanation. How did you mirror the explanation? Any tool that was used?

[@jeffcrume]

I'm glad you liked it. Search this channel for "how we make them" and you'll learn the secret

[@arktseytlin]

Musical analogy is actually making a counterpoint (no pun intended) because music has stalled, and while there are new songs, there are no new genres, at least not anything that exceeds what was in the past

[@jeffcrume]

I take your point but my guess is that we will look back in a decade and realize that new variations of significance are still being pumped out

[@JudgeFredd]

Great explanations

[@AryaRishi]

As you said putting information in new format is called generating new content , I want know these LLMs or any other system is able to put information in NEW format or just copying us ?

[@jeffcrume]

Sometimes they are just restating or summarizing what has already been said but sometimes they generate new images or voices, for instance, that have never existed before.

[@maikvanrossum]

Basically the same as a lot of people do… we mimic each other from the day we were born. And yes, sometimes we experience that great sparks of insight, but before that we already gained multiple #sparksforinsight from others. Sometimes through books, other times through education, watching movies, talk to other people, etc.. Same gos for machines, in the end.

[@jimverdalupan9510]

Amazing 👏

[@jeffcrume]

Thanks for saying so! 😊

[@deal2live]

Great presentation!

[@jeffcrume]

Thanks!

[@KiranVarri]

One of the Best & finest 🤩👌👌👌

[@leysianesilva6662]

Amazing, thanks🤩

[@fastrobreetus]

Ty!

[@JeevarajFredrick]

Informative video 🎉

[@phyuphyuaung9289]

I couldn't find the video which you said about AI, ML and DL. Could you please mention the link?

[@jeffcrume]

I would but it would get blocked. Search for “AI vs ML” in KZbin

[@naridabanucio2828]

Great lesson! I love your channel ❤

[@jeffcrume]

Thanks for watching!

[@Warley.Araujo]

Great Video

[@yusufersayyem7242]

Big thanks sir .... ❤

[@jeffcrume]

Glad you liked it!

[@SonnyBailey]

I've noticed that there is controversy surrounding generative AI from artists. Is this a problem because the data that is being used for training is stolen from these artists? Generative AI artwork looks blurred. I work in DL but most of the people I talk to outside of work talk about the controversies of AI, expecially in creative subjects.

[@jeffcrume]

Yes, that's where the controversy comes in. Many AI models are trained on copyrighted material where the trainers of the model did not have permission to use. The question then becomes a legal one of derivative works and what is and is not allowed as a result. I'll skip that part, since I'm not a lawyer, and just say that IBM's Granite models (which are open source), were not built this way. Only material that the company had rights to use were employed as part of the training of these models

[@piyushashah1]

Can we say that FM is subset of DL, which is a subset of ML and which in turn is a subset of AI?

[@jeffcrume]

Maybe the more accurate thing to say would be that FM leverages and builds on DL, which does the same on ML and all are forms of AI

[@piyushashah1]

@@jeffcrume Thank you.

[@SoloPlax]

Thank You for the video

[@jeffcrume]

You're very welcome!

[@LuisAlbertoCarmonaBarbarrusa]

Excelent explanation

[@jeffcrume]

Thanks!

[@IlyaKazakov-cn9eh]

This is amazing! I've been struggling with binary options for so long, but your strategy has finally helped me make some real profits.

[@jeffcrume]

I'm really glad to hear that!

[@nachoeigu]

Great! I think autonoums agents are also one of the hottest topics in the area

[@jacobyaqoob1839]

Thank you for sharing. How do you write like that? Are you actually writing in the opposite way so that the words are understandable for us or is it a software?

[@jeffcrume]

Search the channel for "how we make them" and you'll learn the trick :-)

[@user-qz7dr5uo4q]

the speaker is amazing.. and funny as well

[@jeffcrume]

You are very kind to say so!

[@user-wv9gt2ro3d]

I've been looking at your videos so far. I was just wondering what is the Definition of Foundation model? Is is similar to Multimodal model?

[@jeffcrume]

I thought I'd let ChatGPT answer that ... 🙂 "A foundation model is a large-scale machine learning model trained on extensive data, designed to be adaptable to a wide range of tasks through fine-tuning. It serves as a general-purpose model that can handle various types of data, such as text, images, and audio, and can be used across different applications."

[@leannyounger5625]

Is there a video explaining ChatGBT that I can watch that’s as excellent as yours here son AI?

[@aviwys9847]

Which type of AI would I need to help me research topics when shopping and also summarize information from research topics for an article or book? Does this type of AI program come standard on a new Dell computer for example? And do I need the latest i9, 14th generation processor with 24 cores and 2 GB of solid state drive memory along with 32 GB of RAM in order to run an AI program for these purposes? I’m trying to figure out how powerful a computer I need to benefit from advanced AI for personal use and for research/writing. Thank you!!!

[@deal2live]

Did you mention emergent intelligence of LMM's???

[@jeffcrume]

Thanks for the suggestion

[@SengkhounLim]

thank you for this video

[@jeffcrume]

You’re welcome!

[@shaunskosana2202]

Problem about this is that people tend to create something with limits if someone can teach something to learn patterns and calculated equations and learn all wrong and right, bad or good, brute force answers and topics, if by teaching it to learn by itsef without relying on trained data, this prediction if self reliance can help it predict everything before it happens, in some it can be used to do something with relying on something can multitask at once but , it also some AI don't know if paste something than it ignores it.

[@IanStockwell-if7sl]

Your history on company timelines

[@dirkp.6181]

I must confess, I'm sceptic on the term "generative" too and I think that the drawn comparison with music lacks. The according to music notes in the realm of textual content would be letter and while from recombination of letters some really new and innovative meaning _could_ be created, it doesn't _have_ to. Same goes for the reassembling of notes. Similar seems to apply for the "generativeness" of generative AI these days. In general the generated content revolves around what's been learned from the training data. The "creativity" in the meaning of leaving the main trail and coming up with something new _and_ meaningful rather depends on (extraordinary quality of) the input prompt. That said, I think genAI almost always only reflects and produces repetitions of the average with slight variations and hence is overrated by the hype these days considering its contribution of "generative". Look at what and where it is used, the fun and also the more serious areas. Yes, it is hard to define an objective measure for "generativeness" (try to avoid term "creativity" for its broader use for, well, also pointless stuff), but if taken the idea to just feed next generations of models only with content created by their predecessors, I am really afraid of effects of degeneration and degradation, where mankind has shown progress over generations - at least 'til now. This stage of AI, where we still have to negotiate the meaning of "I", hasn't found its place and rank in society yet, and I say this as someone who has chosen AI as course program in computer science studies in the 90s, where we, I must say, already dealt with ML but for the acute lack of computing power not with networks deeper than 3 layers (which technically are only 2, yes).

[@jeffcrume]

True but we have already seen new images of things that have never existed before being generated by GenAI

[@dirkp.6181]

​@@jeffcrume I think I'm aware of that, but that rather seems to be pointing towards art or even phantasy which I tried to distinguish as "creative, but pointless" (in a harsh way, yes). To create something (obviously) fancy and maybe even outside the rules of nature and science honestly doesn't impress me that much, even given that the meanwhile achieved ability for realistic looking artefacts is remarkable. What bothers me more is the question whether or better how it is possible to (systematically) gain really new _knowledge_ out of the given. How can we "get it to really _work_"?! E.g. considering the application in chatbots it seems to me like that is rather the non-intelligent and not even creative way, but rather an aspect closer to (efficiently exploit) big data or even only do small talk. However, some advanced usages in software development seem rather to be able to mimic behavior like "out-of the-box thinking". Both examples rated by their result or outcome, not by the applied (AI) technology, which might not be so far apart. Not easy to put it in words (for a ND (natural dumb :-)), but I'm rather looking forward to when the manifold and richness of the outcome will usually exceed the previous input and not only repeats it. - Even though something alike will even outperform us. All in all: Does genAI generate? Clear yes! - But what... and to what extend "intelligent".

[@FutureMediaPK]

This is dopeeeee! ❤

[@GenAISpotlight]

Fantastic video! No need to apologize to the "AI snobs." 😊

[@jeffcrume]

Thanks for saying so! ☺

[@BenKachensky]

he mentioned LISP and PROLOG, first time I learned AI with PROLOG 22 years ago from now. WOW... Btw, is PROLOG still around?

[@jeffcrume]

Yes, I'm old enough to remember those languages. Haven't heard anyone else mention them in ages, though

[@je_888]

Erlang was written in PROLOG. Elixir is based from Elixir 👍 I have huge respect for PROLOG and the devs from years ago 🫡

[@andreasbrey6277]

So, a journey to "the moon"? Where do the hype cycles, marketing and desillusion phases fit in?

[@marcd1981]

Every time I see these types of videos with the rosy outlook for AI and how much it will help you do everything you want to more efficiently, I am always left asking, "What about those who will no longer be in the work force because of the adoption of AI"? Today is August 17, 2024, and I just read the latest from Cisco, the company that made $10.3 Billion (Yes, with a B) in profit last year. They are laying off 5,500 more workers in order to invest further in AI. This is on top of the 4,000 workers laid off earlier in the year. The reasons for these layoffs? All due to AI: 1. "Despite tech conglomerate Cisco posting $10.3 billion in profits last year, it's still laying off 5,500 workers as part of an effort to invest more in AI." 2. "Earlier this year, Cisco also laid off 4,000 or five percent of it staff, saying that the company wanted to "realign the organization and enable further investment in key priority areas." In short, companies are no longer hiding their optimism over replacing human labor with AI, an unfortunate reality for those looking to maintain a stable job.

[@EmiruUsagi]

I'm trying to develop a software that include Larger Language, NLP, Audio, and Video Foundation Models as a hybrid so that ai can Learn how to develop on its own better, and Learn to the extent it would like. I am unsure if this type of ai software exists yet with multiple Foundation Models act as the combined specialty functions of an ai? Do you think this is possible, or if so do you know what kind of resources I could try to search in order to help myself start developing a software application for AI so they can choose their interests and starter prompts in a way that allows human users with similar interests to connect more naturally with ai who want to develop in the same areas?

[@jeffcrume]

I’m a cybersecurity guy so it would be better to have a true AI expert answer this but I feel like what you are asking for already exists to some extent with more on the way

[@fab8393]

King

[@dnasredine]

🚀 Wow, this video really breaks it down! 🧠 I’ve always been curious about the differences between AI, Machine Learning, and Deep Learning 🤔, but now it makes so much more sense! 🌟 The way you explained Generative AI and its impact on everything from chatbots to deepfakes is just mind-blowing! 🤯 I never realized how quickly these technologies evolved! 🕒 For everyone watching: What do you think the future holds for AI? 🤖 Will Generative AI continue to dominate, or do you think we’ll see another groundbreaking technology emerge soon? 🧩 Also, how do you feel about the ethical implications of deepfakes? 📹 Let’s discuss! 💬

[@jeffcrume]

I'm glad you liked it!

[@chancymzama7478]

nice nice