Everyday AI Podcast – An AI and ChatGPT Podcast - Ep 419: Ask me anything AI: Grilling Jordan on everything AI

Episode Date: December 10, 2024

❓ Will AI take my job?  ❓ What AI skills are most important?  ❓ How should my company be planning our LLM strategy?  ❓ Are you an AI bot?  I've talked to hundreds of the smartest people in the world on AI and asked them all of the burning questions.  We're flipping the tables with this episode -- you can put ME on the hot seat.  Get your questions in now, and I'll be answering them live on this special edition of Everyday AI. Newsletter: Sign up for our free daily newsletterMore on this Episode: Episode PageUpcoming Episodes: Check out the upcoming Everyday AI Livestream lineupWebsite: YourEverydayAI.comEmail The Show: [e-mail] with Jordan on LinkedInTopics Covered in This Episode:AI News Rapid-Fire Audience QuestionsAI Fears and ApprehensionsEmail Management with AIEveryday AI's Future PlansAI Use in Podcast PlanningAI Effects on Human InteractionsAI in Email ManagementAI tools Usage and PreferencesEveryday AI Team DynamicsAI Use on the WebAI Free MarketOpenAI Future AnnouncementsAI Trends and ModelsAI Model Testing and PreferencesSora Processing PlanTimestamps:00:00 Sign up for daily newsletter, Meta's nuclear AI.06:12 Technical difficulties prevented planned live Q&A.08:22 Transitioning to Windows to use AI for emails.12:05 Relying on sponsors, podcast not profitable.16:25 Prefers MeetGeek and Otter over Google Gemini.19:19 Ensure guests fill form; use narrowed GPT.24:02 AI content ubiquitous; human connection remains irreplaceable.25:55 AI replacing human interaction may become normal.27:43 Expect more value soon from $200 plan.33:14 Primarily uses 3.5 SONNET; plans testing Nova.34:28 Haven't tried Sora $200 plan yet due to busyness.40:13 Everyday AI: Learn, collaborate, and explore together.40:47 Join live stream, newsletter for AI updates.Keywords:Everyday AI, Jordan Wilson, OpenAI, Google, Microsoft, Meta, nuclear energy, AI projects, quantum computing, Willow chip, Sora video tool, technical difficulties, live stream, audience questions, AI in email management, LinkedIn, AI concerns, AI future, AI accessibility, AI in communication, AI in human experience, AI tools, Chat GPT, CoPilot, AI-free services, Artificial General Intelligence (AGI), AI in entertainment, AI in gaming, custom GPTs, notebook LM.Send Everyday AI and Jordan a text message. (We can't reply back unless you leave contact info) Start Here ▶️Not sure where to start when it comes to AI? Start with our Start Here Series. You can listen to the first drop -- Episode 691 -- or get free access to our Inner Cricle community and all episodes: StartHereSeries.com Also, here's a link to the entire series on a Spotify playlist. 

This is the Everyday AI Show, the everyday podcast where we simplify AI and bring its power to your fingertips. Listen daily for practical advice to boost your career, business, and everyday life. Meet Firefly AI Assistant, now live and Adobe Firefly, the All-in-One creative AI studio. Just describe what you want to create and the assistant handles the rest, orchestrating multi-step workflows across Photoshop, Premiere Express, and more in one conversational interface. You direct the outcome. The assistant accelerates execution. I've been lucky enough to be able to talk to some of the smartest people in AI in the world, right?

From the fast-growing startups like OpenAI to leaders at big tech companies like Microsoft and Nvidia and IBM and also small business, entrepreneurs, Fortune 500 leaders, I've brought them all live to you on this very show. Yes, this is a podcast, but it's also a live stream, but I thought I've been able to let you guys ask questions from hundreds of guests, but never really let you grill me. So that's what we're going to be doing today on a special edition of Everyday AI, kind of a hot take Tuesday, but you all get to grill me. All right, I'm excited for this one. What's going on, You know, my name is Jordan Wilson and welcome to Everyday AI. This is for you.

It's a daily live stream podcast and free daily newsletter, helping everyday people not just learn about AI, but how we can all actually leverage it to grow our companies and our careers. And it starts here, yes, on the live stream slash podcast, but where you really make this go to work for you is on our website, your EverydayAI.com. Go there, sign up for our free daily newsletter. So yeah, we bring you guests on every day except today, right, or most days.

We have a guest on. And we actually write a recap from the most important insights from the guests as well as in our newsletter, bringing you everything you need and you need to stay up to date and to get ahead. So make sure if you haven't already done that, go do that. All right. So before we get in and let you all grill me with your questions, we're going to start as we normally do with the AI news. So first, meta is reportedly.

turning to nuclear energy for its AI power needs. So meta, according to reports, is exploring nuclear energy to power its energy-intensive AI projects, aiming to operationalize nuclear reactors by the early 2030s, according to sources familiar with the discussion. So this move aligns meta with other tech giants like Google, Microsoft, and Amazon, who are also either considering or already making moves on nuclear power to support their AI initiatives. So Despite its reliability and low carbon emissions, nuclear energy remains controversial due to safety concerns. All right.

Next piece of AI news, a huge one. Any other day that we didn't have a big announcement from OpenAI, this would have been our top story. But Google has developed a quantum computing chip named Willow, which can complete tasks in five minutes that would have taken the fastest classical computers, 10 septillion years. I don't even know how many a septillion is.

So the chip is comparable in size to an after-eight mint promises to revolutionize drug development by expediting the experimental phase, according to reports. This breakthrough is part of a series of advances since 2021 in quantum, suggesting that quantum computing could transform research and development of new materials within five years. Willow is noted for being less prone to errors than its predecessors and could significantly enhance artificial intelligence capabilities. So quantum computing leverages the principles that matter can exist in multiple states simultaneously,

enabling vastly larger calculations than classical computers. And obviously, this is pretty big for AI developments, right? Well, it's big for everything, for life, for technology, for scientific discovery, for everything, right? But think of what life was like before computers and the Internet and AI. and how much that's changed, take that times like 20 in terms of what quantum computing could actually do for our everyday lives. All right. And then last but not least, the big piece of AI news for the day, OpenAI has officially

launched SORA. It's new AI video generation tool that allows users to create videos and images in various styles from realistic to abstract. So this marks a significant expansion into the AI video space for OpenAI. and this comes about after 10 months. Yeah, we first heard SORA 10 months ago when Open AI previewed it. So we really had to wait for a long time and we finally got it. So this is part of Open AIs, kind of 12 days of Open AI for the holidays, running daily announcements until December 20th.

So SORA is available to users through different subscription tiers. so only with chat GPT plus the $20 a month plan and chat GPT Pro the $200 a month plan. So if you are on a free plan, you're not going to be able to access this. So you can go to soror.com and there is a credit system. And I'll tell you this, the Chad TPT plus plan, the $20 a month plan that so many of us are on, you're not going to get a lot out of Sora. So if you really want to push it or really use it all, aside from some basic experimentation, you're not going to get that.

So on the chat GPT plus plan, there's always water. You can't use humans. You have to be on the $200 a month plan. So, you know, my initial testing, I did a video on it. So make sure you check out the newsletter today. We'll be linking that. You know, it's impressive. The outputs are good.

Love the user interface. But yeah, in terms of limits right now, super limited unless you are on that $200 a month plan. And even on that, if you are an avid user, a hardcore creative, this is not going to, I don't think, you know, give you what you actually need in terms of limit compared to other programs like Runway or Luma. All right. Let's get into today's show. And, well, a little bit of a bummer.

So I promise, y'all, I wasn't trying to dodge any questions like Allison on LinkedIn was saying. So we actually attempted the live stream this morning. And sure enough, our live streaming platform wouldn't let me go live, right? So that's just how technology works. I don't think I've ever done this in 400 plus episodes where I'm like, all right, we're going to, you know, you guys can grill me live, ask your questions. And sure enough, after 20 minutes, I realized the live stream wasn't working. Well, people told me after three to four minutes and I tried to troubleshoot. So, you know what? If you know of any other easy live stream platforms, it looks like we've been using Stream Yard. So I'm going to have to jump and use something else. So, you know, we always, if you're listening on the podcast, which on this one, well, you definitely are. since the live stream is broken, let me know what your favorite live streaming platform is. We're going to have to make a switch.

All right, but don't worry, I'm still going to let you all grill me because I did copy and paste over all the questions. And I promise you this, some of them were long questions. I didn't even look. All right. Some of them that were short questions, obviously when I'm copying and pacing them, I can see. But I wanted y'all to actually be able to grill me.

So I didn't, you know, some of you I can see left long questions. I didn't read those. I copy and paste them. Put your name in there. So let's get to it. We're going to go rapid fire through all of these questions. It's hot take Tuesday. So I'm going to try to give you my honest, just bold take.

So without further ado, let's get into the questions. And most of these are from LinkedIn. I'm sorry, we had multiple streams on YouTube. In the first stream, I couldn't pull the questions off. So if you were listening on YouTube, I'm sorry about that. Might have missed some of the questions. All right. So Allison and Mark just kind of joking around saying,

Are you just trying to dodge our questions? No. Mark, love this. This made me joke. Does Jordan have laryngitis? Yeah, because I was talking for like three to four minutes before anyone could notice or before I noticed that the audio wasn't working. Woozy said, hey, first question, how do we hear you today?

Well, that's on the podcast, I guess. Fred joking around. How about using your Jordan avatar? Yeah, I could have. I could have. All right. Now let's get into the real questions here. I had to shout those out because some of those made me chuckle.

All right. So Michael asking, do you use AI to manage your emails like auto respond calendar? If so, how? Good question, Michael. So the short answer, and I don't know if anyone else asked about this, but when it comes to just what AI I use, I'm currently in the transition from going from Mac products to Windows. So I literally in front of my face have a brand new, what is this? This is a Microsoft copilot plus Surface PC 7 or something like that. So I've tried inside of Google Gemini to use

it to help with emails. I've tried a handful of other programs. Some work well, some not so much. So the only, if I'm being honest, I don't get a ton of email. I unsubscribe to everything. I suck an email, right? Some of our partners that reach out to me all the time know this. So for longer email threads where I find it hard to keep up, Gemini is okay at that. But that's one of the reasons why I'm switching over. So I've been using the copilot subscription on my Mac, but not all features are available.

Like an Outlook as an example, there is no copilot if you're using it on Mac. whereas in Microsoft Word in Excel, there is co-pilot. So yeah, I'm definitely looking forward to the Gemini email kind of offering leaves a lot to be desired. It kind of stinks, and I've reviewed this on the channel. So right now, I'm using AI a little bit to manage my emails, but not a lot because, number one, not a lot of people email me. And number two, when they do, it's generally a little shorter. But when there's long email threads, big projects, I do.

copy and paste it and kind of strip out any, you know, PII or anything like that. And then I will chat usually with notebook L.M or with Claude. Just does a little bit better with long context windows. Adobe just introduced an entirely new way to create, bringing the power and precision of its creative suite into one conversational experience. Meet Firefly AI Assistant, now live in the Adobe Firefly app, the all-in-one creative AI Studio. Powered by Adobe's creative agent, Firefly AI assistant lets you start with your vision, just describe what you want, and shape the outcome as it takes form with the assistant.

The assistant orchestrates multi-step workflows, drawing on 60 plus prograde tools across Adobe Creative Cloud apps, including Photoshop, Illustrator, Premiere, Lightroom Express, and more to help bring your ideas to life. You can also get started with creative skills, a growing library of pre-built workflows for common creative tasks like batch editing photos, creating mood boards, portrait retouching, and creating social variations. Every step the assistant takes is visible so you can refine, redirect, or take over at any time. You stay in the driver's seat as the creative director.

Adobe Firefly AI assistant now in public beta. See it today at Firefly.adopi.com. All right. A couple more questions from Mike, saying from Michael Peggins. who joined us on YouTube and LinkedIn. So thanks, Michael. And I know you sent me some recommendations on live streaming. So I'll check that out.

So Michael, saying, do you actively use any AI agents at the moment? I test some out. There's new ones every day, right? Am I using anything consistently? No, not right now. Again, with my workflow right now, you know, it's pretty simple. So I have everyday AI, right? That's one of my companies.

We do this podcast, this live stream. You know, we have sponsors, right? We have advertisers like Adobe and Microsoft are our last two advertisers. So, you know, aside from that, I do have another business, Accelerant agency. We've been doing digital strategy for companies for, when did I start Accelerant? About seven or eight years ago. But I've talked about that. We've been winding this down.

So I actually don't have, this might sound weird. Aside from this podcast, which I'm not great at making money on, I don't have a lot of client work per se. So, yeah, not using any AI agents at the moment, aside from testing them out. I do think that will change once I get this PC setup and using Copilot Studio, right? Windows or, sorry, Microsoft 365 Copilot, extremely excited. As most of you know, I was at the Ignite Conference, was invited. out there and just being able to talk with the people actually building copilot studio, right? I had Ray Smith on, who was the VP of AI agents at Microsoft.

I talked to probably at least 10 other people at Microsoft who are responsible for building agents, right, across their entire line of product offerings. And I'm extremely excited to kind of finally get this computer set up. All right. Michael also asking, how many people are on the everyday AI team? Good question. Very small. We use a lot of AI, even though we don't use a lot of AI agents.

We do kind of duct tape a certain AI software together with Zapier to make everything super simple. But right now, there's really two of us full time, have another quarter time contractor and other people helping out on an ongoing basis. But in terms of like, who's actually staff? Like 2.25 of us. All right, Ted. Ted asks, are you using the check? GBT extension in Chrome as a replacement for search. What are the advantages?

Ted, great question. Yes and no. So yesterday, I actually disabled it because at least for me, it's not where I want it to be yet. So for about half of my searches, I loved it, right? The other half hated it. Here's why. So as an example, a lot of times, I would just be putting in a simple,

like, I don't know, a restaurant or a store, right? And I have that habit in my brain, you know, because I use Chrome and Edge. And when you do that, essentially, you get like a local business and you get their address, their phone number, their map, right? So for those type of searches, I was going bonkers by using the chat GPT extension. For the other half, great, right? So some of these more phrase searches and things like that. I know also I can just type in Google and hit tab with the chat GPT extension. But if you're new and you don't know the chat GPT extension, it is great for a lot of things, right?

But essentially, when you use the address bar in Chrome or in Microsoft Edge because it's based on chromium and it works there as well, essentially when you would go into your URL bar and type of search, instead, it automatically just launches that in chat GPT. So like I said, for certain searches, fantastic. right? But if it's something, if it's something very simple, not so good always. So you just have to keep that in mind, you know, but try it out. Try it out. It's definitely worth looking into if you haven't yet. All right. Next couple from Cecilia here. Cecilia is asking which of the meeting apps in the Zoom Microsoft Teams vein is best at incorporating AI. So there's pros and cons to

Zoom. And I've had, you know, multiple people from Zoom on the show. One thing that I don't like about Zoom's AI is I don't know if it actually records a full transcript, which is what I like most of the time. I haven't used Google's, if I'm being honest, right? Like, I'm not huge fan of Google Gemini across the Google workspace. For whatever reason, it doesn't work that well. So I've tried it in Google Meet, don't like it. And I haven't tried it in Microsoft Teams, because like I just said, still in the kind of transition phase. But for meetings, I like using either Meet Geek, which is something I've talked about on this show before, as well as Otter. So at least for me right now, probably until I fully integrate myself into Microsoft and Teams,

which I'm excited for because Microsoft did just at Ignite announce some new kind of teams agents. So a facilitator agent and also like a translation agent like you can literally speak another language in real time, which is nutty. All right. Thanks for that question. Cecilia. Also asking what is your favorite model to use for voice transcription? Great question, Cecilia. So I do use a lot of other.

programs for this, right? And sometimes these different programs, you can choose the transcription model, right? So I use Otter. I use Otter AI to transcribe audio. Also, cast magic. We use that on a literally a daily basis for this podcast, but I also run some things locally. So as an example, Mac Whisper is a cool program that I use on my computer sometimes to transcribe audio locally there. Super fast, super cheap. You just plug in your API key. But for the most part, I believe that OpenAI's Whisper model is probably the best and most accurate, although I did see some rumblings recently online. I'm trying to remember. Let's see if I can look it up here. I think it was fireworks AI. I think it was one of the models that I looked at and I'm like, wait, that's pretty good

in terms of like transcription time. But yeah, Cecilia, for the most part, I'm using Whisper from OpenAI. Great model, easy to work with. All right, Samuel asking, I believe this one was on YouTube, says, do you have any custom GPs that you use regularly? Absolutely, Samuel. So I do a lot of them for planning this podcast, right? So one guests submit to the show.

So there's a couple process. There's a couple steps in our process, right? So, hey, if you're even listening to this, let me just throw this out there. If you are listening to this and you're like, wow, my CEO would be great for this show or our, you know, CMO. I love to get them on. You can go pitch, right? We hide this on our website, but you can literally just go to your everyday AI.com slash pitch, right? Pitch your guests.

We get a lot of guests that pitch to come on the show. We don't take a lot of them on, but we do take some of them on if there's a good story. so make sure you fill that form. Anyways, once we get a guest, once we accept a guest on the show, because we, we don't want people just coming on and selling you guys, right? We want to get good guests who can really bring value. But anyways, once they do, there's a little form that they fill out. And then essentially, I take the results of that form.

And sometimes it's a lot. It's a lot of information. And I plug all of that into a custom GPT that I've built that helps me. I have a couple different ones. And yes, I could use. one GPT that does all of these things. But I'll tell you all, as I tell everyone else, I think GPTs work best when you get them a very narrow scope.

Don't try to have them do everything. So I have a series of, I believe, three different GPs that I sometimes use for podcast planning. The one that I probably think is the best or the one that I use the most, well, it's helping me just come up with titles, right? I know that sounds dumb. But a podcast title means a lot. Right. So usually we have the guests submit one or two. I'll have an idea for one or two. And then I essentially copy and paste all this information. And a lot of times, right, and I have chat

Chebt go do some research, right, based on the information. So it's not just like, here's a bunch of text, go give me podcast titles. It's like, no, I have custom instructions in there that, you know, instructed to browse with Bing, you know, find trending topics in that field. Because sometimes I might not know, right? Like had a pretty great guest a couple of weeks ago on quantum computing. I don't know quantum computing. It went out there. This GPT, the custom GPT found some trends and it gave me some good suggestions, right?

Still, human in the loop, right? I control what information I put in there on that GPT. It kicks things out. I look at them. I'm like, yep, this is good, right? But I give it good instructions, right? I say it needs to be a clickworthy title. It needs to have a, you know, one of the main keywords in the front end of the title, right?

Because when you're scrolling through on Spotify, you don't even, you don't always get the whole title. So simple things like that, as well as doing additional research on topics that I don't know. So I have another one that helps me research. Again, this is just the first steps of research. I'm still the human, right? I still go through and read all this. Yeah. So those are just two GPTs that I use pretty frequently.

We also use projects. I use gems sometimes in Google, Google's version. I don't think they're that great, but that was like three months ago when they first came out. I'll probably give them another run, as well as we use a lot of other tools. You know, desktop tools like Bolt. We use, what else do we use? Po. I like using chat hub.

So, you know, some of these where you can build essentially the equivalent of a custom GPT in these other platforms. So, yeah, I use the concept of these. custom GBT's very often. Cecilia, what are the chances that the AI community will stop using acronyms and speak clearly so they can be understood by everyone? Yeah, asking about just accessibility. So I'll say probably not a lot, right? Like as an example, if I, every time I wanted to say reinforcement learning from human

feedback, that's tough. That's a mouthful, right? So a lot of times I just spit out RLHF, or RLHF, right? Or Rag, right? If I'm talking about Rag, like I always try to first reference, right? Maybe for my old journalism days, I always like to say retrievalogue many generation, at least once. But then I'm like Rag, right, in your Rag pipeline, right? So I don't know.

I think for the most part, and sometimes people don't even know the acronym. So maybe even Cecilia, maybe I should do a better job at making sure I do this. All right, Guillermo, former guest from way back in the. the day. Quite a few questions here. So I'm going to try to answer them quickly one by one. So he says, as AI advances and generates everything flawlessly, perfect deep fakes, digital clones, optimized content, do you think we'll see a growing market for authentically human experiences? G, abs so frigging-lutely, man. That's why I do this, right? I saw this three years ago before I even decided to start a show, right? Because I even at that time, I'm like, yo,

I know pretty soon a deep fake Jordan is going to be just as good if not better, but I can interact with you all live in real time. I stutter sometimes, right? I say the wrong words. I do think that there will be something in the future where kind of like it's like, you know, if something's, think of like going to a farmer market, right? And you're like, oh, this is local. This is handmade, right? This is handcrafted. I do think that AI generated content is going to be so commonplace in a year or two, you're literally going to see people, you know, marketing themselves or their products as not using AI, right? That's why I love doing this thing live. And yes, I could just use an AI avatar or I could clone my voice on 11 labs and

it's really good. And I could just not do this podcast, right? I can just have AI do it. I do think one thing that AI will not replace is human connection, even though you all can't tell right now. In theory, right? In theory, this could be my 11 labs clone. It's not. It's actually me. That's why I take questions from the live audience, and obviously this is the day where things don't work. G, also asking where AI can create anything,

could the real thing, quote unquote, become a luxury? Yes, exactly what I'm talking about. I do think in the future there will be a premium on human-made content. This is multiple part questions here from G. So also saying, will we, the pre-AI generations, start craving unfiltered raw human moments, Like live concerts, yes. Oh, absolutely.

Saying like AI free weekends, yes. It's, if I'm being honest, I think it's going to get overwhelming. I think I think AI's infiltration. And yeah, you might be like saying like, oh, okay, Jordan, AI infiltration. You're the one pushing this. Yeah, I am. But I'm also trying to help you all understand it and help us all understand it. Because I think it's actually going to be much more consequential than we actually think.

I don't think AI and generative AI and large language models are going to be this. thing that continue to just live on the internet and we have to go seek it out. No, AI content, AI entertainment, AI interactions are going to find you whether you want them or not, right? And you're not going to have a choice. I was thinking the other day, I'm like, there's going to be times, I think, probably in 2025, where you're going to be interfacing with a company. And the only option you will have is AI until there's a law, right? You might go on their website, it's a chatbot, it's AI. You fill out a contact form or an email in AI response. You call them. It's an AI voice agent. I literally think that there's going to be some companies in the

near future where you will not be able to interact with a human, which I think is terrible, right? Like, I want y'all to know this about me. Just because I talk about AI and I enjoy it and all these things, I don't think that it needs to replace all of us humans, right? I think in some situations, talking to an AI will be much better than talking to a representative and waiting on hold for 45 minutes and talking to someone in a call center and you can't hear them. because it's, you know, their keyboards loud and there are five coworkers, like, I have bad hearing. I can't, I can't hear that, right? I'd much rather, in some instances, talk to a human.

But I do think that there will be almost a luxury market for that. And then Guillermo also asking, is there a future market for products and experiences explicitly branded as free of AI? Absolutely. I've already seen that a little bit out. I think with some creatives, some photographers, you know, they just say, you know, not using AI. But yeah, I do think it's going to, people are going to lean into it heavily. All right. Philip, Philip's got five questions.

So I got to go quick here. All right, Philip, what's your take on the chat GPT pro plan? Well, $200 a month. It's pricey. But I've been saying all along, literally, I said this a year or two ago. I said, I wouldn't be surprised. Before there was all these official reporting from the information, I said, I wouldn't be surprised as someday there's a multiple $100 or multiple thousand dollar tier.

of Open AI, right? As they get toward agentic AI, you can't have that for $20 a month, right? You literally can't. So what do I think of the pro plan right now? Well, I think that we're going to see a lot more value to it in the coming days, right? So we're only three days now into OpenAI's 12 announcements. So we're going to see nine more announcements from Open AI, and I'm guessing at least half of them will have features that are only available or mostly available on that $200 a month tier plan. So it's too early to tell right now, right? So right now, the only differentiators are the 01 Pro model, having full, no limits on the GPT

4O, and then better limits on SORA. So right now, I think $200 is a little steep, but I'm pretty sure we're going to see some things over the next nine days of announcements that will even more justify that $200 a month price tag. So second question, what are the top five trends expected to shape the world of AI in 2025? Oh, gosh. I don't know if I can think of five on the fly. Like I told you guys, I didn't like re-research these questions beforehand.

I wanted to give you my honest reaction. I don't know if I can label off five, Philip. I do know we're going to be doing another show on like 25 predictions for 2025. So I want to give this some thought. I know right away AI, like customize AI entertainment. It's going to be everywhere, right? I think we're going to get in 2025 custom AI TV shows, shorts, websites, games, right? Literally being able to generate live personalized games in real time.

So I will say agentic AI. I'll say voice AI. I'll say real world or real time entertainment. saw some things InVidia was working on that they haven't released yet. It's going to be pretty wild. But we'll do a full show on that soon. Philip also asking which three large language models are expected to stand out the most in the coming year? Great question.

I mean, obviously chat chabit, right? They're so far ahead. People are always like, especially like, you know, hardcore AI people like online on Twitter, on Reddit. Everyone's like, oh, Claude, Cod, Codonet, right? I think it's like one of those things right now. It's like you feel trendy by saying, oh, son it. you know, Sonnet 3-5, 35 News, Sonnet 3-6, it's better, right? It's better in a few things.

The only thing that I use Claude for right now, if I'm being honest, is artifacts, right? So that's the kind of the split screen or dual pain ability right now for Claude to render something in real time. But between ChatGPT's Canvas mode, which does that for writing, and I do think that we're going to see over the next nine days, Chat Chepti released their version, could even be today. I do think that they're going to be able to, they will release their version of artifacts. So number one, chat, CBT, number two, I mean, it has to be co-pilot, even though that's technically not a large language model. It has to be co-pilot. And then I'll say, you know what?

I'm going to say one of the Chinese models. All right. Maybe Quinn, very impressive. I think they just got internet access today. I think China is going to become a leader in the L.A. L-LM game in 2025. They lacked behind from 2022 until the second quarter of 2024, and now they're a dominant force. So that'll be my answer. Philip also asking, when will we achieve AGI and what are the key principles guiding its development?

Good question. So I'll say this, depending on the definition, we're technically already there. All right. A staff member on Twitter from open, sorry, an open A. employee on Twitter said he thinks they already have achieved it, right? If you look at the definitions of artificial general intelligence, and if you're new here, that's essentially when a single AI system is essentially smarter than the smartest humans at almost all tasks.

If you look at the definition from 15 years ago, we've 100% achieved AI, but the definitions are always changing, right? Literally, go on to archive.org, right? The internet archives. Go look at definitions. I did this for a show, so you can go listen to that show. Go look at definitions of AGI from 2008, 2012, 2015. We've achieved it, right?

According to those ones, but according to the AGI definition of 2024, everyone has a different definition, right? It's, I think, fortunately, unfortunately, Open AI is the one that decides it because of their partnership with Microsoft and their HGI clause, which we've been covering, right? So, I mean, a lot of people see it, right? And especially with these agentic reasoners and these new models like 01, the QWQ model from Quinn, these reasoning models. When you combine a reasoning model with agentic, right, agentic capabilities, I'm like, yo, I don't know. Is this AGI?

Probably. Maybe. Maybe not. I don't know. So it's more of who is defining what AGI means now. All right. Fred, asking, with so many changes happening, it's sometime difficult. to know which model you think is best. What are your five favorite models today, front end,

and for what purposes? So, yeah, like, if I'm being honest, I really only use frontier models, right? So chat GPT or the GPT4O models, MATA 3-2, been testing 3-3 a little bit, Anthropic Sonnet 35. I don't really use anything else from Anthropic, from, you know, the Claude models, aside from the new 35 Sonnet or 35-Sonet. two, three six on it, whatever you want to call it. And then for the most part, I test other models. You know, the Quinn models, so I'll go into Hugging Face, test out models. But for the most part, at least if we're talking about kind of like quote unquote text-based

models, I'm not using a lot of them, right? I do have to try out Nova a little bit more from Amazon. But Fred, for the most part, you know, it's like, you know, in Gemini. I like to use Gemini in the AI studio because on the front end of Gemini, it's using an old model. So, yeah, I stick to the basics. I know that makes me, like, pretty basic and boring, but that's the reality. And then, okay, so for what purposes?

So I love using Notebook L.M, which is powered by Gemini for just factual information, right, as a grounded model. I like using Claude, kind of like I mentioned, for its artifacts feature, being able to render things in real time. I like using ChatGBTBT for just about everything else. I like using Copilot on the web, even though I know, co-pilot is not its own model. It uses, you know, GPD 40. Tara, have you tried SORA on the $200 plan? I was curious to see what the difference in speed for processing SORA on the $200. So I have not yet. I'm still on the $20 plan. Life's been a little busy over the last couple of days. I got actually doing some slight home renovations in case over the last couple of days you've been hearing banging in the background because I normally record this show at home. so I haven't actually had time.

Hopefully I will this week to fully try out the $200 plan. And I know they have been kind of restricting signups as an example for SORA. So I know I've got to get that going sooner rather than later. But like I said, Tara, did do the $20, the normal chat TBT plus SORA. I do know you get more credits. I think you do get some like priority access, you know, when things get busy. But I haven't done the $200 plan yet. Tara asking, what are your thoughts?

about the AI comments and DMs people are receiving on LinkedIn. I have mixed feelings, right? Part of it, right, LinkedIn is encouraging you to use AI, right? It has that right post with AI. I'm sure that they're going to, at some point, roll out that same thing for comments. I wouldn't be surprised, but yeah, there's dozens of Chrome extensions and services that, you know, just write AI comments. And some of them do. it automatically without you even browsing or without you even approve it. You just connect your LinkedIn, which is against LinkedIn's terms of services. And it just automatically comments things with AI. So not a huge fan of it. If I'm being honest, I don't think there's any way to necessarily

stop it. All right. And I think we got to all the questions. It's our last question. Yeah, we had about 20 questions. I went through them all. Looks like last question here. And again, y'all, I'm sorry the live stream got this kicked off because I'm sure we would have got some more questions in from you. So maybe we'll do it. next year, right? I feel I'm going to curse myself if I say, oh, we'll do it again next week and it's going to not work again. All right, but Ben, last question, love this one. So saying, as someone who's seen way more behind the scenes than most of us, what's the one thing keeping the experts up at night? Great question, Ben. I have to ponder about that, right?

Because obviously, I talk to a lot of people. Some of them are quote unquote on the record, right here on the show. A lot of our guests I talked to before. before and after the show, things that maybe they didn't want to say in their official position. When I go to conferences, as an example, I talk to people all the time, a lot of people. I'll say this, the unknown. I think that's what is keeping the experts up at night. Because even the experts don't necessarily know, right? They don't.

I mean, it depends on what area of AI, right? Artificial intelligence has been around for many decades. I get it. But if we're talking about the new way of generative AI and large language models and their development, you know, transformer models, you know, reasoning models, right? Even the experts, even the ones building them. If you listen to, you know, go read their long essays, their research papers, they go on podcasts, right?

Even the experts, there's an element of this technology that they don't understand. So I'll say that is the thing keeping most experts, aside from implementation and measuring ROI, right, like the boring answers. But otherwise, I'll say it's the unknown. No one knows, not even the people building the technology. No one truly knows what's next, right? They know what they're working on next. They know what the, you know, consumers are demanding, but no one necessarily knows what's actually happening next, right? This is, you know, I think one of the first times. At least in my, you know, I'm going to be nice to myself and say 30 plus years of life. I can't say that too much longer. I think most technology, right, it's the humans telling

the machines what's possible. Now it's humans challenging machines to create a new possible. which is scary, it's thought-provoking, right? It makes us all rethink intelligence. It makes us wonder what's next, right? But I've talked to literally dozens of people, some on the show, some not, that have said that exact same thing. The thing that keeps them up in night is, hey, machines and AI are creating our future. And we have to live in it.

We have to make the most out of it. All right. Nothing like ending on a super deep question like that from Ben. So all right, y'all. Thank you. Thank you, especially to all of our live stream people that, you know, stuck around for 20 minutes as I was trying to get our tech figured out this morning. So hopefully we get it figured out here soon.

And hey, for the podcast, people, thank you for listening to this as well. I know it's a little different one. I wanted to give, right? I'm always so very appreciative of the people who attend live, whether it's on LinkedIn or YouTube or, you know, some people join on Twitter, right? So if you are a avid podcast listener, we have great guests all the time. And you can actually, I know it's 7.30 a.m. Central Standard time. So, you know, it's 8.30 if you're on the East Coast.

It's 5.30 if you're in California. It might not be the best time for you. But you can always ask. even after the fact. We always put the link to the LinkedIn post, well, except for today, in our podcast description as well. So that's one thing, you know, getting back to Giromo's question about kind of authenticity and humans' role in AI. I want everyday AI to be that place where us humans that are just trying to figure it out, right? No one here is, is trying to, you know, say that they're the leading expert in anything. This is the place that

you go to learn about AI and how we can all work together. and leverage it. So if you are an avid podcast listener, please consider joining this on the live stream sometimes. Stop by, stay high, ask a question. So thank you all for sticking around for this interesting episode. If you haven't already, please go to your everyday AI.com. Sign up for the free daily newsletter. We will be recapping. Yeah, I guess we'll even do a short recap on today's episode. But we recap it every other day, as well as in our newsletter, we tell you the latest news, fresh finds from all across the internet, upcoming shows, everything you need to know to be the smartest person in AI at your company. Thanks for tuning in. Hope to see you tomorrow and every

day for more, everyday AI. Thanks y'all. Meet Firefly AI assistant. Now live in Adobe Firefly, the Allman One Creative AI Studio. Just describe what you want to create in your own words and the assistant handles the rest, orchestrating multi-step workflows across Adobe Creative Cloud apps, including Photoshop, Premiere Express, and more in one conversational interface. You direct the outcome while the assistant accelerates execution. Stand control with the ability to step in and refine at any time. See it today at firefly.adobie.com. And that's a wrap for today's edition of Everyday AI.

Thanks for joining us. If you enjoyed this episode, please subscribe and leave us a rating. It helps keep us going. For a little more AI magic, visit Your EverydayAI.com. and sign up to our daily newsletter so you don't get left behind. Go break some barriers and we'll see you next time.