Subscribe
Share
Share
Embed
Marijn came back from MVP Summit in Seattle, but can't say anything, because NDA! The topic changes in what AI can do for you, and why people are the secret ingredient for AI success! They end with an American whisky "Straight Triticale Distillarium".
Creators and Guests
Host
What is Office 365 Distilled?
This podcast is presented by Marijn and Steve; two experienced Microsoft 365 and SharePoint consultants. In each podcast you will get a smile or two with a chunk of interesting and practical help in Microsoft 365. Then you get introduced you to what we hope is a Whisky you have not tasted before.
Hey. Hey. Tritikale. Tritikale. Yes.
Speaker 1:Pretty pretty Tritikale. Tritikale. Yes. That's what you bought back from Bloody Seattle for... To drink tonight.
Speaker 1:Tritikale?
Speaker 2:Yes. I went to the MVP summit, and all I got was a Tritikale.
Speaker 1:Tritikale. I'm really quite interested to see what this thing pops out like. So, yeah, Distilled Arium, I will cover it a little bit later. But a triticale, it's a little bit like this podcast.
Speaker 2:This this is pretty triticale as well. Yes. Yes. So,
Speaker 1:apparently, triticale is a single grain that has characteristics of wheat and rye. Mhmm. So it's a... Let's say, oh, it's a genetic creation. We're gonna drink this and grow horns.
Speaker 1:Oh, boy. Okay.
Speaker 2:We'll see.
Speaker 1:We'll see. Blend. I know. You've tasted it, so I'm guessing you're impressed There's with more than just a shot gone from there. Yep.
Speaker 2:Yep. Yep. Yep. I've been... I drank the naughty water a fair bit of two.
Speaker 2:Yes. Since I got back.
Speaker 1:Nice. Yes. You said you had 12 whiskies when you left this place, and then you missed out my spirit bar because she was too gone.
Speaker 2:Yes. Yes. Yeah. Well, it happens. I mean, you're walking around in the city and you happen to run into a distillery and then well, what do you do?
Speaker 2:You need to go and taste it.
Speaker 1:That's true. So you have been traveling a lot and you still haven't finished this month yet or next month as well? No, this month now.
Speaker 2:Yeah. Yeah. It was it got busy.
Speaker 1:Yeah. So in four, five weeks
Speaker 2:I flew to, Seattle for MVP summit, came back. Next day, flew to Sarajevo for the Adriatic Tech Summit. And then in two weeks' time, I'm gonna fly back to the East for Orlando for the big m three sixty five conference. That's gonna be a hoot.
Speaker 1:It will be. A very, very hoot hoot. Who's the star of the show? Do you know yet?
Speaker 2:Oh, deeper, of course.
Speaker 1:That's the boring technology. I mean, last PowerPoint one, they had one of the big female singers I can't remember the name of now, but she was pretty awesome, filled the whole stadium. Who was it? Who's sang?
Speaker 2:Oh, I I don't think they're gonna have, like, a party like this. Why? Wasn't they Christina Aguilera
Speaker 1:or something? That's right. Exactly. Why would they not?
Speaker 2:I don't know. They haven't done this Forum three sixty five conference.
Speaker 1:No. That's true.
Speaker 2:That was the power platform conference. They've got more money.
Speaker 1:They've got more money.
Speaker 2:They've
Speaker 1:got They've more using AI and power platform.
Speaker 2:Absolutely.
Speaker 1:Yes. Yes. So then you're gonna go off to Orlando, but then you're gonna come back. And a few days later, you're gonna go to... Oh, no.
Speaker 1:That was where you
Speaker 2:talking about Hamburg.
Speaker 1:Then the next
Speaker 2:one is the, collab summit in Cologne, where we're all going.
Speaker 1:Yeah. Busy. Busy. Busy.
Speaker 2:Yeah. And then in June, I'll be at Experts Life UK first. Then two week or two weeks later, I will be at
Speaker 1:Commsfurst. Commsfurst,
Speaker 2:and then we'll travel to Hamburg, and then I'll travel to the European platform conference. And then it's time for summer. And it's time to get some work done because a lot of customers will be very angry.
Speaker 1:Yeah. Well, part of the fun. Yeah. Part of fun. You're enjoying it?
Speaker 2:I am. It's weird because MVP summit was just two weeks ago, but it already feels like months and months ago. So crazy.
Speaker 1:Alright. So I'm gonna sit there and... Because I never got to Seattle. I did not to go this year.
Speaker 2:Yeah. Mhmm.
Speaker 1:And so I just left you to it. Yeah. You had some very, very late nights. Honestly, the photos that kept popping up with all the people that I really wish I could see as well.
Speaker 2:I thought... We had a few a few good nights. Yeah, of course, we had the things you always do, like Joey's, karaoke at Palmer's, obviously.
Speaker 1:Jesus, yeah.
Speaker 2:Thank you, Amanda, for putting the most difficult song for me and you to sing All
Speaker 1:the little old village bar. I just love it.
Speaker 2:Yes. It's a lovely little divers bar. Yeah. And yeah, exactly. Yeah.
Speaker 1:Yeah. Cool.
Speaker 2:So, yeah, no, it was it was good. It was fun. Yeah. It was fun. Neat.
Speaker 2:Yeah. So I guess
Speaker 1:I should question you what you learned.
Speaker 2:Can't say anything about that.
Speaker 1:You can. There's some things you can and some things you can't.
Speaker 2:There were some really good sessions, so obviously I was gonna question you. Okay.
Speaker 1:Otherwise, you just pick the bits that you like.
Speaker 2:Yeah, of course.
Speaker 1:So I'm gonna add some really difficult questions, I suppose, really. So, the all time best whisky you tasted?
Speaker 2:Oh, the the best whisky I tasted that week has to be that sherry bomb.
Speaker 1:Oh, yes. You said that.
Speaker 2:Cask from the Copperworks Distillery.
Speaker 1:Same distillery as this one? Yeah. Yeah. Yeah. Oh, no.
Speaker 1:Copperworks. Copperworks.
Speaker 2:So they said, this is what the Octamore is for pitted whiskey. This is what this is for sherry cask whiskey.
Speaker 1:But you said it was too much.
Speaker 2:Yeah. They were not joking.
Speaker 1:But you still liked
Speaker 2:it. But I wanted
Speaker 1:to buy a bottle,
Speaker 2:and I couldn't because they didn't have any left. So otherwise, I would have bought it. Wow. Yeah. No.
Speaker 1:That's cool. So let's just talk about AI generally. I'm guessing most of the week was AI stuff.
Speaker 2:There was there was a little mention here and there. Yep.
Speaker 1:I can hear the sarcasm already. Yeah. So did you get a feel for change in the Microsoft strategy around AI or just same speed of growth or new areas of interest? Let's just talk about where we are and then you can work out what you can and can't say. I mean, I think everybody knows that we're under non disclosure for a lot of this stuff, so we're kind of not down.
Speaker 1:But what I wanted to do is get a feel for what it was like, what other people were thinking about AI and that kind of stuff. So did you get a feel that it's still new or that it's now bedded in?
Speaker 2:It's it's definitely not new anymore. But what I do have the feeling or still the feeling of is that they're still trying to wrap their head around where AI could be an impactful difference.
Speaker 1:Okay. That's an interesting term.
Speaker 2:So I've seen AI being used in some areas that I wouldn't have thought about, but that actually make a lot of sense.
Speaker 1:Yeah. I do strike me that to get the real value out of it, you kind of have to think really out of the box. Yes. I mean, we're we're talking earlier about some workshops and stuff that I want to kind of put together, but I'm thinking about using AI and Copilot, but you end up using it in the same place that you would do other forms of, you know, authentic. I do a session where I basically try and use AI to give you a score for collaboration.
Speaker 1:Yeah. So it goes off and it starts looking at everything you've done and then comes back and said, yeah, you could have done better. Yeah. He's always said you could have done better. Let me suggest how you could do better.
Speaker 1:No, AI will start.
Speaker 2:That was awesome. Let's see how you can do better.
Speaker 1:Yeah. That's what I've just said. Do you not listen to me when we do this podcast? No. No.
Speaker 1:Okay. That's alright. Just open your mouth and see what comes out. Yeah. Yeah.
Speaker 1:Yeah. Alright. No. That's cool. So yeah, no, I find the same thing.
Speaker 1:We were also talking about today that user stories in an agile environment. Yes. To be able to write a prompt that kind of checks and looks for improvement on your sprint. So which items did you get finished? How many story points did you do?
Speaker 1:Was their priority correct? Those kinds of things are not what automation would do for you, but it would actually at least try and help you to see things in a...
Speaker 2:That would that would be like your scrum master agent or something.
Speaker 1:Yeah, maybe. I think
Speaker 2:I think it would definitely be doable to spin up an agent like that.
Speaker 1:The problem with people say agent to me, I just think workflow. Because they're all saying it's gonna do this for you and, you know.
Speaker 2:Well, no. Basically, what it needs to have is a trigger that it runs at a certain time, like, for example, at the end of your sprint or a day after your sprint has ended. And then it will proactively come and tell you how it went.
Speaker 1:No. Thought you're I think you're wrong. I think you're wrong. I think you need to walk in the room and AI says, okay, you guys. Listen up.
Speaker 1:You failed. Yeah. Alright. You missed this. You missed that.
Speaker 1:You missed the...
Speaker 2:Drill drill master drill sergeant agent. Yes.
Speaker 1:Sergeant AI. Yeah.
Speaker 2:I can see that.
Speaker 1:Yeah. Yeah. But it's driven by the meeting starting. It's driven by a time, by an appointment.
Speaker 2:No. No. Absolutely. It could... Yeah.
Speaker 2:It it... That that could definitely work. And now with with the new co work, rolling out, I mean, I'm I'm sure we are barely scratching the surface of what this could be.
Speaker 1:Yes. So you heard about some pretty exciting stuff coming down the line?
Speaker 2:Yes. But also, I mean, co work is out. You can already play with it. This is on on the frontier model. And, of course, our good friend from Israel, Amy Diamond.
Speaker 1:Was he there?
Speaker 2:No. No. He wasn't there. Oh. But he he has been playing around with co work.
Speaker 2:He already put some videos out
Speaker 1:Oh, nice.
Speaker 2:Which I definitely need to see. But, I mean, he's definitely playing with some interesting use cases on on how to how to work with this. But... No. I've I've I've already played around with the AI for SharePoint, the the the Knowledge.
Speaker 2:Thing where it can create pages and it can create your libraries and your things. Definitely,
Speaker 1:They it's can't just automation things. Just... You could have done that in workflow. You could have do that with a, you know...
Speaker 2:You can't really do that in workflow, but you could... I mean,
Speaker 1:you gotta build the workflow the same you gotta build the agent, just that they built it for you. There's no reason why No.
Speaker 2:It's it's... I mean, the the the the of this is that you can, without any IT knowledge, talk in your language to the agent, say what you want, and the thing will create it for you. And it could be a page or it could be a library with metadata. It could be whatever.
Speaker 1:But I come back to your comment. They're still trying to find out where AI is going to add value. Yes. All that's doing is the same as you clicking a few buttons and typing a few titles in.
Speaker 2:It's just doing pretty quickly. Learn. It's like like, for example, one of the companies I work together with, they're using HubSpot. I've never used HubSpot in my life. So now I need to, let's say, classify leads or like the...
Speaker 2:I've... Let's say I went to a conference and I've got some leads and now I need to put them in the system. I've never used HubSpot before. So what if there's an agent where I could just say, I talked with this guy from this company that looked very promising and he asked me questions like, okay. Do you think they're ready in six months?
Speaker 2:And that will build up the whole lead chanting without me having to learn how to do these things in HubSpot.
Speaker 1:So we're gonna go to work, sit down at a desk, computer screen will stick on, Holly will have a head, ask you a few questions, and then you're gonna have a coffee. Yeah. Work will be so desperately exciting.
Speaker 2:Yeah. But but I mean, that's that's that's that's the whole premise. That's the whole idea.
Speaker 1:That work will be so exciting.
Speaker 2:No, no, that we... That there's a bunch of complexities taken out of the things that we need to do.
Speaker 1:I think you chose a bad example. Okay. Because I think when you're doing sales and sales leads, there's a lot of emotive stuff in there as well.
Speaker 2:True.
Speaker 1:So yeah, how do you feel they went on with it rather than and...
Speaker 2:Okay, yeah, sure.
Speaker 1:But I get what you're saying. Do what you're
Speaker 2:for example, I was having a meeting earlier with a customer and the HR director doesn't know anything about SharePoint. She knows she wants a SharePoint, an HR bit on the internet, but she doesn't know how to create pages. And frankly she doesn't probably doesn't really care. And what the SharePoint agent does or the AI in SharePoint does is she can just describe what she wants to the agent and it will create a page, make it beautiful, put all the relevant bits and bobs in there without her knowing about web parts and whatever.
Speaker 1:I get that. So that's all still seems like it hasn't changed then in the last two years. May have got better at doing it.
Speaker 2:It's it's... It definitely... I agree with you on that one. It definitely got way better. But, basically, when it just came out, my expectations were pretty high.
Speaker 2:When I finally... Yeah. Exactly. So when I finally got to play with it, I... So I was thinking this could be like my intern.
Speaker 2:But when I was playing with it, I figured out that it was actually a three year old toddler
Speaker 1:which is drawing on the walls
Speaker 2:instead of making a beautiful design. I now feel that the toddler has been growing up, and it is now 12, 14 year olds, let's say.
Speaker 1:So now it's about to start to get to its
Speaker 2:teenage years. It can do some things where I will go, that's interesting.
Speaker 1:You're right. I think when it gets to 14, it will get to the point where it can do so much more Yeah. But doesn't know not to.
Speaker 2:That's also true. That is also true. A few weeks ago, I was, delivering a training course around building an intranet, and I said, okay. One of the things that I found very important is that you give your intranet a good name. And I said, let's use AI to come up with a good name.
Speaker 2:So we had the idea that our company was a coffee roasting business.
Speaker 1:That's original.
Speaker 2:I know. Super original.
Speaker 1:Starbucks and Seattle. There's quite some coincidence there.
Speaker 2:No. No. No. I I was just using that because that gave me, like, good visuals to play and stuff like that. And so, basically, in the...
Speaker 2:Before when I asked AI, like, come up with 10 good 10 good names for my company, it would just say coffee hub, coffee place, coffee Internet, coffee whatever. Like super basic stupid 10 answers. And now when I asked it, it came back with some really, really, really cool ideas where I was like, if I had an actual coffee company,
Speaker 1:this would be the name.
Speaker 2:This was really, really cool. And I don't have my phone with me. Otherwise, I would look them up, but they were they were really, really... Yeah.
Speaker 1:Memorable names.
Speaker 2:Yeah. And I was like, interesting.
Speaker 1:It has
Speaker 2:grown up. And now I think with the co work, which is finally the thing that I think... So basically when this came out and I was hearing voices certain voices from people in The Netherlands that do a lot with Co Pilot saying I've been talking to Co Pilot throughout all my car journey and he does all these things for me and then I was trying that out and it just does jack shit for me. But now with the co work, I now see myself sitting in the car talking to copilot saying, you know what? What's on my agenda today?
Speaker 2:Oh, this and this and this. Oh, you know what? Send an email to this guy asking about this topic or that he could explain this before this meeting. And he will send those emails. It will do all those things.
Speaker 2:Now we finally get to that stage where I was hoping that we would be two years ago. Now for me, AI finally makes sense because it can do things instead of just giving an answer.
Speaker 1:Okay. I don't know.
Speaker 2:He said not convinced.
Speaker 1:So I wonder whether they can tell that it was you that wrote it or whether it It doesn't matter.
Speaker 2:It's fine. It doesn't matter. The only thing is that I have an idea and I want to communicate that idea to that person. And I'm letting somebody else write it down.
Speaker 1:Interestingly enough, I'm guessing that, hackers and virus generators have now got some new direction to go in.
Speaker 2:That's definitely something else, yeah.
Speaker 1:Oh well. Yeah. Well, let me know when you send your first email to the wrong person.
Speaker 2:Oh, yeah. Yeah. That that will probably happen way sooner than I... Than we all think, but, yeah.
Speaker 1:Mhmm. And this is part of a standard Co Pilot license?
Speaker 2:This is part of the Copilot license. You need to be in Frontier and you need to turn on the Anthropic model processor thing. That's There is a GDPR thing because the Anthropic is running on US servers and is not running on European servers. So before you do that, check if that's okay within the company. Check with security or whatever to
Speaker 1:Yeah. But that will change eventually. It's nothing
Speaker 2:Oh, no. That will change. But as as of now, it's it's not it's not GDPR compliant or or not compliant because the processing happens in The US. But for me, for my one person company, I I don't I don't care. So...
Speaker 2:Yeah. I mean, I'm not handling any trade secrets or whatever.
Speaker 1:So your job on your drive home tonight is to send me an email from your car.
Speaker 2:Deal. I will I will try to do that. Yes.
Speaker 1:That'd be cool.
Speaker 2:Yes.
Speaker 1:And then also in that email, you should describe the image to be created for the podcast so it sends it to me and then I can just publish the podcast. Mhmm.
Speaker 2:Yeah. Yeah. That's that's the challenge. I'll I'll try to... Yes.
Speaker 1:I'm gonna try to
Speaker 2:do that.
Speaker 1:That's fun. I have to say, just don't have time for the time being. It's been on my list as well, but...
Speaker 2:Same.
Speaker 1:We'll do the same. Yeah. So the whole frontier thing, course, is turning into its own little environment, isn't it? Its own little space, I'm guessing, not just a word anymore. From a Microsoft perspective, I think they've been really quite smart with the way that they have kind of marketed and positioned this.
Speaker 1:I think that this whole Frontier thing was a bit of a marketing thing to start off with, but it's now starting to create its own identity and its own sort of halo. We're going to innovate in AI and
Speaker 2:Yeah. This is Sure. At the Adriatic Tech Summit in Sarajevo, I went to a super interesting session from Mike Madarani.
Speaker 1:K.
Speaker 2:Don't know if you know him. I've seen him around in various places. But I've never been to one of his sessions, and his session was how to create an advanced agent through Copilot Studio. And he started off with, like, basic first steps and then just went deeper and deeper and deeper and deeper and deeper. And I don't know if you've ever built a chatbot back in the day with the, what was the power platform thing named before it was Copilot Studio.
Speaker 2:It was
Speaker 1:something something. Chats.
Speaker 2:Yeah. I can't remember. But, basically, building an advanced chatbot is... Or an an an agent chat, whatever, how you... Whatever you wanna call it, still has a lot of those old chatbot functionalities in there.
Speaker 1:Of course. People like the AI. They know what it looks like. It's on websites
Speaker 2:or Yeah. But I mean, if this person asks this thing, then you need to ask about that thing, and then you need to go there, and then you need to grab this for that answer.
Speaker 1:Whereas you expected it to be kind of fully automated and now work it out for yourself.
Speaker 2:Because, basically, I've created a few agents. And for me, creating an agent is, here's a library with documents. Publish. The agent is born.
Speaker 1:Yes.
Speaker 2:And it's now finished. Basically The six instructions. He did was he create... He started with that, but then he was like, okay. But now if people are asking this, then this will be their intent and you need to go that and then you need to grab this information from that source and do this.
Speaker 2:So that was that was really cool.
Speaker 1:And do you think he was saving any time doing that from what he used to do? Because I mean a lot of these things are still building the foundation for what will happen in the future. I mean, I get it, this is what it does now, but they're really looking for where to go.
Speaker 2:So the idea is that you will build a lot of agents that all do very... Something very, very specific. Yes. So for example, I've tried this out with creating a a YouTube... Shit.
Speaker 2:What's it called? A YouTube manager for me that would come up with names or that would analyze my channel and do things like that. So basically, I just said, this is my channel published. Agent is now live. And just said in the description, you're a YouTube manager, whatever, handle this.
Speaker 2:And, basically, what I now learned is that I could, with a few tweaks, could nail it down much more and that it would be much more powerful in doing that one thing that it needs to do.
Speaker 1:No. And it's the same with things like meeting templates as well. Yeah. So you can actually do a... What does a meeting do?
Speaker 1:I need an agenda, and I need you to do a transcript.
Speaker 2:So, basically, your daily stand ups, for example, yeah, you could have a daily stand up manager. And if you have a one on one, you can have a meeting template for a one on one agent meeting thing.
Speaker 1:But the value is when you put time and thought into it. Yes. Because otherwise, it's just a simple meeting with the building blocks of gender
Speaker 2:Oh, no.
Speaker 1:Transcription. No.
Speaker 2:Exactly. That's what I'm saying. Like, if you build the agent with a number of those specific things in mind, then it can be way more powerful.
Speaker 1:Yeah. Which still takes time. One thing I'm becoming more aware of is AI is not going to save you that much time, but it is going to allow you to do things of a higher quality. It could create your email. You end up having to go in and edit the email, or you end up having to go look at it, or you end up having to go change it, or you say it's not quite the tone that I want.
Speaker 1:So, I mean, like you said earlier, it's gonna take away the repetitive stuff. Yeah. Mhmm. But you do at some point go, you know, I could have just written this myself straight out of my head with a higher Yeah. Level of version.
Speaker 2:So I've got that, for example, I was delivering a training a few weeks ago, and they also wanted Outlook training. And more specifically, they wanted more information on how to make yourself out of office, how to share your calendar, how to... I don't know. Two more things. So I just said to Copilot, you know what?
Speaker 2:I've already got everything for OneDrive and Teams. I just need need you to create me an Outlook bit, including these questions for end users that barely know anything about Outlook. And it create me six, seven slides. Super boring. So I basically had to remove them and start all over again.
Speaker 2:Yeah.
Speaker 1:Yeah. Because if you try and get them to change them, it starts to argue with them and give you instructions.
Speaker 2:Yeah.
Speaker 1:No. But that's okay. I mean, this thing is relatively new. I mean, one or maybe version 1.2 or 1.5.
Speaker 2:Yeah. And and for for some reasoning things, it... It's it's pretty good. I mean, I let it create some documents and some ideas and some planning things and some offerings and things like that. And it it it does save me some time because I don't have to write all the extra bits myself, and I can just tweak it whenever I want.
Speaker 2:But, also, it does add some of the bits to it like paragraphs and structure that makes it a more higher quality document.
Speaker 1:No, but I don't disagree with that. But it also, if you're not careful, can lose the mistakes that you should normally have picked up yourself.
Speaker 2:Yeah, exactly. Yeah. I read an interesting article a few months ago. I don't know if we talked about this on the podcast, that in the future, there will be two types of employees. One type of employee that will give AI work, and one type of employee that will check the work of the AI.
Speaker 1:That's interesting.
Speaker 2:Yeah. It...
Speaker 1:I... There was a Harvard manager is back.
Speaker 2:A Harvard Business Review, article. Yeah. That was that was pretty interesting. Yeah. So that...
Speaker 2:That's... That was some food for thought.
Speaker 1:Yeah. Always strikes me that it's an ideal agent to check and an ideal agent to create. Because an agent can check spelling quite easily in grammar, and it can also feed feed you back what they... The the reader will feel from it.
Speaker 2:Yeah. So that's something I like. For example, like, could say, make it more professional, make it more playful, make it make it funnier. Yep. Yeah.
Speaker 2:Those things I like. But I mean, yeah, that that makes it also like a higher quality because if I wanna create a session description and I wanna make it a little tongue in cheek, I can just ask it to do that because I'm creating these in English and my English is not always that good that I can find the exact words that I wanna say that would make this sentence like pop.
Speaker 1:What am still waiting for a lot of this stuff is how I can make AI be me.
Speaker 2:Yeah.
Speaker 1:So with my sense of humor, with my style of writing. So for example, I start That's gonna most
Speaker 2:take a lot of nuclear power plants.
Speaker 1:Yeah. But you end up reading the email that's been created or something, you go, yeah, but that's not me. It's not the way I would do. Yeah. And I have tried several times for so many hours in fact to try and say, hey look, this is me, this is the way I write, this is what I do, this is my humor.
Speaker 2:I had some spelling mistakes in there so they'll know it's me.
Speaker 1:Exactly, yes. To an extent, there's nothing wrong with that. You want to be...
Speaker 2:Always write organization with an s, not with a z.
Speaker 1:Yeah. They're both acceptable in the English language.
Speaker 2:Yeah. Yeah.
Speaker 1:But I but I do get it. You should be able to say English or American. I I think that there's there was a film some years ago. It was called Dylan or some single name, and it was basically about a robot boy Mhmm. That got lost and so then got adopted by a family.
Speaker 1:It was a bit shitty.
Speaker 2:Yes. I think I've seen it. Yes.
Speaker 1:But there was a scene when they eventually found him again and they started reviewing his life. And so, you know, they were playing back the shit he'd done. One of the things he did was he was playing baseball. So this is a robot with, you know, eyes, it can slow the ball down and everything else. And he said, but you missed that shot.
Speaker 1:And he went, yeah. He said, I found I was more accepted by the humans by not being perfect. Yeah, exactly. And basically, he continued that way to getting the balance right. Yeah.
Speaker 1:There's certainly... So AI needs to be imperfect. Oh, it already is. Wow. I am finding more and more that if I have a conversation with my AI about something that I find she's wrong And then you go back and say, hey, no, that's not right.
Speaker 1:And you know you're right, well spotted. Yeah. And I don't know how often people will just accept it for what it is.
Speaker 2:Yeah, but that's that second job is like check the output of the AI. See if that tracks if that's right.
Speaker 1:Yeah, but I don't know how many people do AI in team. So if you go to a consulting company and you write a 25 page document that goes to the client, there's already a checker that will take that and go through it and make sure it's in the style of the company and everything else. And that's what their job is. So in some organisations they already have that, but in many organisations nowadays what they want to do is, hey let's put the quality in at the beginning, Yeah. Because then we can save this and that's been manufacturing for years.
Speaker 1:Used to be a guy that would, at the end of the production line at Ford Motor Company in Daggenen, the largest production plant in the world at the time, he would get in the car and he would drive around the corner, flick on all the switches, drive around there, go and park it and that was it. Quality control done with his checklist. And there was, imagine one car comes off every two minutes and it takes twenty five minutes. That's 12 people per shift that have to Do quality control. The cars.
Speaker 1:And so they went, no. Stop. Look. If we do a better job putting them in and maybe even do a check at the time for the end of each one, then we don't need the guys at the end. Yeah.
Speaker 1:That was If add the a
Speaker 2:few seconds here and there then... Yeah.
Speaker 1:But of course, they could then audit those people to make sure they were doing the good job as need be. But you still ended up with a Friday afternoon car.
Speaker 2:Yeah, yeah, yeah.
Speaker 1:Whatever way around, anything after 04:00 on a Friday guaranteed to be more problematic.
Speaker 2:Yeah, yeah, yeah.
Speaker 1:And the other thing is I find with AI is it's still not there yet to add the context in. You really have to put in the where I'm coming from when I ask this question.
Speaker 2:Yeah. And I think I think that is something that will be solved when we let AI just more into our lives. So for example, with the smart glasses, where AI is just constantly looking together with you. And then you go into a whiskey store and you grab a bottle and then you can ask you about what it is. Yeah.
Speaker 2:And then it has all the context. Oh, well, much more of the context.
Speaker 1:One hopes. I still cannot, because I was thinking of doing this the other day but decided it wasn't gonna work. I still cannot take photographs of all my whiskies and expect it to be put into a database in... Within AI. Mhmm.
Speaker 2:Yeah. And basically what you want is you just want to just make a video of them. Just go with your with your video over all the bottles and then just say, now put them in a database instead of
Speaker 1:images. Them. Yeah. Yeah. Memory is getting better.
Speaker 1:Yeah. Am seeing Yeah. But I know I'm just feeling it. Mean, it's just remembering what I was doing. I was just telling you I was doing my electrical stuff and started talking about...
Speaker 1:The bills, yeah. A new water boiler. And today I was doing something else and she said, yeah, but this is the solution you want. This suits you more because you are X and Y and Z. So the way I think, way I've, you know, the way I've, so she was able to, she is able to start to remember a lot more than she used to do.
Speaker 1:That's more content.
Speaker 2:Okay.
Speaker 1:So that's quite cool.
Speaker 2:I've never really played with that aspect. I've never really taught my co pilot who I am, what I do.
Speaker 1:Picks up. Depends what conversations you have. She knows I masturbate every Monday afternoon, so she never bothers me. Yeah. So Okay.
Speaker 1:That would probably be a different day for you.
Speaker 2:So co work will now move all your meetings away from the Monday afternoon.
Speaker 1:To the point where the screen comes off said, piss off you wanker. Monday afternoon it's time, go. Guys that was a bit Steve's
Speaker 2:personal time. I've added it in your diary for you. I'll automatically put you on do not disturb in Teams.
Speaker 1:Yeah. Dear, well then basically, Oi, I can hear you've started early.
Speaker 2:Yes. Yeah. Oh, well.
Speaker 1:Yep. That's that's another area. We talk about context, but when's the AI going to get to the point where it's listening and and actually looking all the time around you or you in the office?
Speaker 2:Well, heard
Speaker 1:this frightening that story.
Speaker 2:This thing about... Was it Ragnar, I think? The German... Tall German guy that also had these new AI or Facebook glasses, but they have got, like, a red light on when they're recording or something.
Speaker 1:So people know
Speaker 2:turned on or is it Google whatever? Yeah. So people know that they're recorded, but it's AI whatever. How... Or was it?
Speaker 2:I I heard I it from I can't remember who. That that was actually really a problem that people were not talking to them and and, like, getting out of the way, like, nope. Nope. We're not doing this.
Speaker 1:Yeah. Yes. It's, be interesting when it's accepted or not accepted. Mhmm. Or knowing how often you sit down in your chair and then people trusting that the company is not going to use that to decide when I get paid and when I don't get paid.
Speaker 2:Yeah, exactly. Yeah. Yeah. It's like the whole recall memory function in Windows 11. Like, hey, but you've been browsing for a new tea kettle for half an hour.
Speaker 2:I'm not gonna pay you for that. Yeah. Yeah. Mhmm. Interesting.
Speaker 2:Mhmm. Oh, well, we'll see. I mean, AI is going to be used in our work in places we've not seen in common or at least where I've not seen in common. But it's also growing up more, so it will finally be able to do much more things.
Speaker 1:If it's allowed.
Speaker 2:If it's allowed.
Speaker 1:That's the thing. Twenty years time, of course, nobody will care one way or the other, and it'll be just there doing it and recording everything and anything and True. All that kind of stuff.
Speaker 2:I I've was listening to a podcast this morning, and there was this guy that was that was very, very on... Very onto everything. And he was... They were basically talking about space travel. Yep.
Speaker 2:And the whole SpaceX thing that SpaceX is now shooting more rockets into the sky than any other thing. And basically, in a month, they're shooting up more rockets than there there are days. Something like that. And he was like, you know what? It's gonna be funny.
Speaker 2:In twenty years time, people are gonna say, remember this Tesla that they once started out as a as a car company? Yeah. How funny was that?
Speaker 1:And now he's got his own home of Mars.
Speaker 2:Yeah. Exactly. Yeah. He owns Mars. Yeah.
Speaker 2:Exactly. So, basically, it's the same thing with AI. I mean, in a few years' time, we're gonna... I mean, even now, we are laughing like, oh, remember Clippy?
Speaker 1:Yes. We're still...
Speaker 2:We're we're laughing about, oh, remember Clippy? But back in the... Okay. It was pretty
Speaker 1:1990.
Speaker 2:Yeah. It was pretty, not helpful, but it it was a thing back then. But, I mean, in in a few years' time, we're gonna look back and we're gonna say, like, oh, remember that this Copilot thing could also only give you, like, weird fucked up answers? Yeah. And now we will control everything.
Speaker 1:Or, and we had this conversation today around the desk, fucked up answers become the norm. Maybe? And they're accepted.
Speaker 2:Yeah.
Speaker 1:And so they're How missing right a big
Speaker 2:do they need to be? I mean, we are getting consultants for a lot of things that we do in the company. Yes. We are getting builders in. We're getting architects in.
Speaker 2:We're getting people that put windows in. They fuck up as well.
Speaker 1:True. So... That don't make it right though.
Speaker 2:That doesn't make it right.
Speaker 1:Look at my front It's got a it's got a it's a beautiful white front door.
Speaker 2:Yes.
Speaker 1:With a brown bar down the middle of it because they put the wrong stuff on the back of that panel. So they're supposed to be coming to change it.
Speaker 2:At some point.
Speaker 1:No. I think it's now booked for later this month, but we'll see what actually happens. Yeah. So it'll probably be now be white on the back, but instead of being grey on the front, it'll be, no, it'll be back to grey on the front Some instead bald, of yeah.
Speaker 2:Oh wow, yeah.
Speaker 1:But in theory, you know, you take a photograph, send the picture off and it just says, oh okay, I need to change this, and it does it. Yeah. Instead of somebody having to remember the steps to actually make sure that you end up with the right. Yeah. I do think there's a lot more integration that needs to be done.
Speaker 1:Software's got
Speaker 2:to change Yes. Completely at this point in
Speaker 1:Yes. So there's there's so many things. And so it's not going to be as quick and as easy as we think it might be.
Speaker 2:Oh, it's not. No. No. No. It it won't be easy, and it won't be quick.
Speaker 2:But it will gradually get there until at one point we...
Speaker 1:Forgot what it used to
Speaker 2:be like. Exactly.
Speaker 1:To an extent, I don't disagree, but, you've got an adoption session coming up around AI. And I was having conversations with how quick and easy it is to generate websites. So I need to do a new website, I want to focus on my speaking. So Nick said, oh that's easy.
Speaker 2:I've got this, yeah.
Speaker 1:Yeah, he put this thing together and five minutes later he's come back and goes, this is your new website. That's okay and fine. When you start doing that internally though with your people and then you start to think, okay, how do I do the change and adoption when this thing every five minutes is creating new features, new functionality and everything else and you don't think about all those kinds of things and then also the changes and all that kind of stuff. So it is changing the way we approach stuff, the stability of our AI, of our IT systems is going to fall over, degrade quite significantly, unless people take a reasonable approach and people won't. I can do this quickly now, why shouldn't I?
Speaker 1:Well, because there are some things that move slower than you do.
Speaker 2:It can be good or it can be quick, you can choose.
Speaker 1:Yeah. Well, that's a quality thing, but I think I think it... We can change and you put up with it or you can change and you leave. Yeah. So, yeah, it's just...
Speaker 2:What a bleak story. What a bleak outlook on life.
Speaker 1:Yeah. It's a interesting thought, isn't it? Oh. But we all know that AI has moved incredibly quickly.
Speaker 2:Yes. Probably too quickly. Perhaps.
Speaker 1:I will
Speaker 2:There are rumors that... What was it? Anthropic has an full running AGI self aware thing somewhere. There... There's rumors.
Speaker 2:But... Yeah.
Speaker 1:Really?
Speaker 2:I don't know.
Speaker 1:Who knows?
Speaker 2:Yeah. We'll see.
Speaker 1:How? Yeah. Exactly. Yeah. Yeah.
Speaker 1:Yeah. Yeah. Yeah. It is bound to happen. I mean
Speaker 2:It is bound to happen. Yes. We're all gonna be batteries at some point. Yes.
Speaker 1:Yes. Be good film now, actually.
Speaker 2:Absolutely. Yeah. So we'll see. Well, let's have fun while Yes. We're still Let's go to
Speaker 1:all that. I know. A cigar. That was a great podcast that you did, but I've improved it no end. And I've changed the voices to ladies because people like it better.
Speaker 2:Exactly, yeah. I went to the doctors this morning to do my yearly cholesterol checkup. I don't need to do that next year anymore. Fuck it.
Speaker 1:That's true. One of the things I do like very much is when you get those results sheet back, is I feed those into AI.
Speaker 2:Yeah, so I should do that because normally I just call my doctor and I'll just say, you know what, how was it? And he'll say, it's good and okay, done. See you next year.
Speaker 1:No. Just... I just photograph... PDF them, photographs, photograph them, said, okay. How is this?
Speaker 1:Yeah. What do I need to do?
Speaker 2:Yeah. And They will say that's pretty good for a 20 year old lady.
Speaker 1:Exactly. A bit like my heating. That's not bad for a heating system. It's only making one shower and a few bowls of water every day.
Speaker 2:That was interesting. So I'll tell you what I did then. Actually, this is Well, the I think you need to tell the full story
Speaker 1:for the Yeah. People And then I guess it'll be time to taste the whiskey. Yes.
Speaker 2:Know what? While you're doing the story, I'll already pour the whiskey.
Speaker 1:Okay, that's fine. So I use AI for just about anything and everything, because say I'm interested in finding out what it can do and can't do. So I have a new house that I've been working on, so I took printed off all my power bills, electricity and gas, and fed them into AI and said, okay, tell me about my power consumption. And so we had this conversation about power consumption, but it was completely under the impression that I had a gas central heating system purely to do with the amount of money I was paying for my gas. And I went, that's not possible because I have a completely electric heating system.
Speaker 1:And no no no, well in that case you're being ripped off with your gas or somebody's stealing it or it's leaking or something.
Speaker 2:Does it take into account that there's a war in Iran and that the prices are high?
Speaker 1:Actually, I mean I was thinking the same thing but yes, I was wanting to find out before the prices go up. Okay. So we sat there talking about it and basically because I'm using an old central heating boiler, when it turns on I think it goes... It's not another one. It's another new one.
Speaker 1:Yeah. Okay. So then I said, okay, so what should we replace it with? So she went, well actually, it's interesting. We've got a couple of...
Speaker 1:I think the small electric one will probably be the way to go. Yeah. And started talking about the different brands. I started asking about the different styles and where this great
Speaker 2:company Drop the gas thing?
Speaker 1:Yeah, I can then cancel the gas, I don't have to pay for the delivery of it, I can just tell Fluvius to Fluvov and then they'll be done. In America you won't understand that joke, but Fluvius is our gas supplier in It's the government organization that runs power. So we started having this conversation and that was cool. So then I said, okay, actually find me which model I should buy. Okay, well we now have these choices, these choices, these choices and I'm suggesting that you really go down for this one.
Speaker 1:So rather than having an instant on, she wants me to have a small tank in there and all that kind of stuff and that would be... I don't need them to do... She worked out what power requirements I needed, what to find a fuse box I needed, she did all of that stuff. It was really like a good conversation. Alright, well let's assume that I go for this and this.
Speaker 1:Now compare whether my gas prices and the electricity for this running it would work. So she went, yeah, it'll pay for itself in seven months. Wow. So I went, okay, that's cool. So where should we go by?
Speaker 1:How much would this be? Well, it'll be about €800. What if it's I buy it in Poland? Oh, that's a good point, really good thinking. So let's go off and find out.
Speaker 1:Says, well, do a deep dive, take your time. So she went off doing her thing and I went to have a coffee and I came back. She went, okay, we have a number of choices here. So this is the model you wanted and it's available in this country at this price, this country at this price, this country at that price and this is what the mileage was, the delivery cost was. Okay, she said, but I really don't think you should take the cheapest one because I'm getting a vibe, that was her words, I'm getting a vibe that they may not have as much information as you think so I recommend this one.
Speaker 1:So I went, okay, let's do that. So went, so I took it. So I clicked on the link, went looking for the right model, I did compared the SKU numbers, I did Belgian ones, made sure. I said, What about the warranty? She went and checked, made sure the warranty was valid in Belgium if I bought it from there.
Speaker 1:So that was good. So as I'm filling out the form, I saw the box that said, Discount code. I went, Oh, is there any discount codes for this? She went, oh, don't know. She went, she came back, she went, I've got five.
Speaker 1:The best one to use is this one, but in theory, is the only one available if you fill out a subscription. So I put that code in and it gave me another €15 off.
Speaker 2:Okay.
Speaker 1:It was amazing. It was absolutely blown away. Great conversation. It was about an hour. I mean I think by the time we did all the stuff, about an hour, so I just ordered it.
Speaker 1:Okay. I'm actually thinking it's going to be delivered tomorrow when I'm not here but I've got to work that out later. Okay. But so yeah, so now I have electric heater installed, I had all the instructions on how to install it, not that I needed them but yeah. Yeah.
Speaker 1:It was really really a very cool way of actually using AI for something practical. I never do Google searches anymore.
Speaker 2:I do a Google search, but I know that the Gemini answer is gonna appear first, and that's nine out of 10 the thing that I'm looking for. So I just do a Google search just to get to the gymnast.
Speaker 1:I don't bother. I just pop open Copilot, and I... Sorry. I pop open That's a GPT.
Speaker 2:But Yeah. But it's so
Speaker 1:slow. What?
Speaker 2:Gemini is so super fast.
Speaker 1:Yeah. And how accurate is it? Well, for
Speaker 2:the things I've used
Speaker 1:Put the same prompt in. Do the timing and find out what you get back for Gemini and what you get back at CheapUDP and just to compare the amount of information.
Speaker 2:In all honesty, I've been using Copilot in Edge browser and that just Yeah. Yeah. Like twenty seconds before it's there that I can actually...
Speaker 1:I do all my AI stuff in Google. The thing about the timing is that I want a more quality answer and Google uses mostly its own index, so it's search index, so you're just going get the same results back. It doesn't really matter, you've got to choose the one that works. And I often put a prompt in and then I go and do something else. It's the kind of thing that I would
Speaker 2:It's go in and out me clicking on the Copilot symbol and then having to wait for twenty seconds. Yep. And then I can start typing in. But part of I get a result very quickly, but I mean, before it's there. Mean
Speaker 1:Part of that is setting up the guide rails.
Speaker 2:Yeah.
Speaker 1:It's working it out and stuff but it's everybody's got to use the tool, you know I've used Co Pilot from more or less day one. South Coast Summit, I think was where I first saw it in line and away you go. Oh well. Yeah, but that was it, it was an interesting exercise. So this one I've actually taken all the way through to the end result.
Speaker 1:I ordered it from a company, as I said, think it was either Poland or somewhere in that area. And yeah, it cost me €30 to have it delivered over here and it was €820 here. Guess what price I got it from?
Speaker 2:600 something?
Speaker 1:Including delivery cost, $4.25. Nice. Same model, exactly the same SKU.
Speaker 2:Yeah. Okay. That's that's nice.
Speaker 1:It is. It worked very, very Well well
Speaker 2:done, sir.
Speaker 1:Alright. So tell me, my friend, you were gonna go to to Sam's my... Is it Sam's The 1,000 bottle Bottle is here.
Speaker 2:Sam's Shop. Yes.
Speaker 1:Something or other.
Speaker 2:I can't can't think of
Speaker 1:the name. I had the name earlier.
Speaker 2:No. So basically, I came off the monorail. So from the Space Needle back to the city center, there's monorail.
Speaker 1:And I was place.
Speaker 2:Sorry? Yeah. To to bike place market. Yeah. And, basically, I was either gonna go with Paul Hunt and Dan Toft to...
Speaker 2:So Paul Hunt is a woodturner, and he was gonna visit some supplier up north. So I be going get
Speaker 1:out of the city.
Speaker 2:Yeah. With a woodturning shop, or I could go with two Finnish guys and have beers on the side of the lake in the sunshine. Yeah. So I chose the easy option. True.
Speaker 1:And so was the woodturning interesting? No.
Speaker 2:No. No. So I went with Vesco and Mika. We had some beers, and then we decided to go for lunch at that place.
Speaker 1:Yep. And then
Speaker 2:And we walked next to the waterfront up to that place, and we we saw a distillery, Copperworks Distillery. So we hopped in, and there was a super cool dude with a nice beard, a nice weird moustache, telling us everything that we needed to know. We had two flights of whiskey that we just just shared with the three of us. Good stories, really good whiskey. So, yeah, that was what we did.
Speaker 2:But this one is not from there.
Speaker 1:No. I know.
Speaker 2:This one is basically from Trader Joe's because I was trying to get some candy for my daughter and some local whiskey, and this is a local Seattle whiskey. It's called Distilled.
Speaker 1:Distillarium. Yep. And it is a
Speaker 2:Triticale straight straight Triticale whiskey 84 proof.
Speaker 1:So this is a a craft distillery. So it's small batch stuff, very small batch, with this slightly mad scientist energy. It is how it describes or she describes it. But this triticale is an interesting grain mix. It's not actually a blend of grains, it's a hybrid grain.
Speaker 1:So it's wheat and rye, and it's basically a single grain that behaves both like a wheat and both like a rye. Nice. So there you go. It's a bit smoother, has a bit more character than a wheat whiskey. But what's interesting is there's no corn in it.
Speaker 1:Good. Maybe there is. Let's just check because American needs to have corn. Mhmm. And no corn at all.
Speaker 1:Okay. So it's not a bourbon because it's not got 51¢
Speaker 2:It's a whiskey. On It's not a bourbon. Yeah.
Speaker 1:Yeah. Oh, it says got a blunt fervid because I don't like any fluff. It's more honest and grain driven. Let's taste the distillate properly, but it exposes a bit underpowered at 42%.
Speaker 2:Yeah. And that is something I learned as well. The number that it's proof, if you divide that by two, you get the alcohol percentage. More or less. So this is 84% proof.
Speaker 2:That's 42% alcohol.
Speaker 1:Okay.
Speaker 2:I learned that from the guy at Copper Works.
Speaker 1:Not entirely certain because a 100% proof is not 50% strength.
Speaker 2:Is it? No.
Speaker 1:No. But maybe it's just the way that number rates. Okay. Maybe. I don't know.
Speaker 1:Maybe I'm wrong, but I'm fairly sure. I love the color.
Speaker 2:Color is beautiful.
Speaker 1:Do you know a drink called Iron Brew? Yes. This is Iron Brew. It's exactly the same color as Iron Brew. Orange, copper.
Speaker 1:Yes.
Speaker 2:Yeah. It's it's definitely a beautiful color. Yeah.
Speaker 1:So did you buy it blind? You'd never tasted
Speaker 2:it before? Taste it. No. I was
Speaker 1:Oh, nice.
Speaker 2:I was in and I saw a local Seattle whiskey, and I thought, you know what? Let's let's get that. It was either that or it was the Trader Joe's Highland Scottish whiskey. And I was like, nah.
Speaker 1:Nah. Fuck
Speaker 2:that. No. I have some interesting... Yeah. So this is batch 14, bottle number 90.
Speaker 1:Yep. It's interesting. It's got a very rye focused nose, but not overly rye.
Speaker 2:On the nose it has, but then on the taste that disappears. And it's sugary, but in a sweet way, not in a bourbony way. Lots of vanilla and candy. Get candy No? And
Speaker 1:Not the first time. I love the mouthfeel. Yeah. But it's not it's not oily. It's actually quite weedy in terms of its volume.
Speaker 1:It's... Yeah. But the complexity is bet... More than I thought it would be.
Speaker 2:Yes. Yeah.
Speaker 1:It's almost as if it can't work out whether it's a rye or a whiskey. Mhmm. Mhmm. Yeah. But But it's smooth.
Speaker 1:It do get... It is very smooth.
Speaker 2:It's super smooth.
Speaker 1:And orangey. I get orange after about twenty seconds on the back of the finish.
Speaker 2:Like some burnt orange. Yeah. Some...
Speaker 1:Yeah. Yeah. Yeah. That's really nice. I do like whiskies that come with that burst...
Speaker 1:That toasted orangey.
Speaker 2:Yes. Me too. Yeah. So that that Jaffa cake. Yeah.
Speaker 2:Yeah. Yeah. Yeah. Do.
Speaker 1:Nice choice, my friend.
Speaker 2:Yeah. I was very happy with this. Yeah. I opened it up at at home and, yeah, immediately So how many
Speaker 1:bottles did you bring back?
Speaker 2:Just one.
Speaker 1:You wasted time. That was a waste time.
Speaker 2:I know.
Speaker 1:But I was checking back again.
Speaker 2:All kinds of stuff. I went to... Because last year when we went, I had this beautiful square bottle that we drank at the house, that we almost finished at the house, basically. And that was a port cask. And I really want to have another bottle of that because that was so
Speaker 1:good. And
Speaker 2:I couldn't find anywhere. They only had, like, default bourbon or American cask or something like that. And that was, like, $90. Pretty expensive for what it... What I thought it was.
Speaker 2:And also, I didn't see any interesting whiskey. I mean, we had the the tasting at Copper Works, and then the two bottles I kind of wanted to buy, they were not available anymore. Yeah.
Speaker 1:It makes no sense to me at all, but it's a craft. Oh, Copperworks. No. But you you would think I'm tasting it, so I sell you the whiskey.
Speaker 2:Yeah. Exactly. But they didn't have any bottles to sell. They only had a few to taste. And, yeah, so I was like, I know.
Speaker 2:But I was kind of almost tempted to go to Westland and to see what they had opened.
Speaker 1:Yeah. What they had left after it was all stolen. Yeah. Yeah. Exactly.
Speaker 1:Yeah. Yeah. But you never got a chance. You was only on there for a quick trip in, trip out. Yeah.
Speaker 1:Exactly. Yeah. So Howl the Moon. Next trip is Howl at the Moon.
Speaker 2:Yes. Yes. I've been I've been I've been looking at that. Yeah.
Speaker 1:I can imagine. And
Speaker 2:then the whiskey bar, the whiskey that we went to. Oh, yes. That cigar place on the other side of the street.
Speaker 1:Yes. Envious.
Speaker 2:So I'm meeting up with some friends, and one of my friends is also a cigar smoker. Oh, okay. Cool. I'll take her there.
Speaker 1:The San Fran... The Frank Sinatra. Yeah. That First time we tasted
Speaker 2:it. Yeah. Beautiful bottle.
Speaker 1:But this is interesting. What does Sandra think?
Speaker 2:She doesn't like this. She doesn't like bourbon and and things bourbon? No. No. She doesn't like American or rye or
Speaker 1:Oh, okay.
Speaker 2:Things like that. So... No. Mhmm. No.
Speaker 1:Yeah. Very, very nice.
Speaker 2:Yeah. I think so too. I'm I'm happy with this shot. I can see this bottle disappear pretty quickly.
Speaker 1:It is drinkable. No doubt about it. It's very But rye whiskies tend to be.
Speaker 2:Yeah. I I like rye because it's a little different. Yeah. And it gives you a little, yeah, hairy, grainy, like... Yes.
Speaker 1:Hairy. You hairy. Yeah. Got it. Basically.
Speaker 1:I understand. Alright. Pretty cool, my friend. Well, thank you for coming over. I know that we're out of sync because of all your travel plans and things.
Speaker 1:Yeah. I know. But this is good. A bit of an AI chat.
Speaker 2:We had a... Yeah. Yeah. It it... I get the feeling that you're more somber about the AI future than me.
Speaker 2:And I think that used to be the other way around at some point.
Speaker 1:I did. Yes, it did. I'm still really looking for...
Speaker 2:Use cases.
Speaker 1:Not Yeah, maybe. But no, because I use it every single day. Yeah. Mean, I I But...
Speaker 2:And I use it once every three weeks or something.
Speaker 1:I think the problem is that I I kind of deal with infrastructure nowadays. Mhmm. Okay. That's that's what I tend to do. So things like stability and managing risk is is a key thing.
Speaker 1:It's it's one of my big attributes. Mhmm. And AI is Is not one of those. Risk. Yeah.
Speaker 1:Because of its unknown ness. And I think that's the other thing. I hate unknowns. Yeah. I work by not remembering, but by working stuff out.
Speaker 1:So, had some great conversations with Nick T. Thanks Nick T. I had a good afternoon today where we were talking about AI and how it actually works. Yeah. And he was talking about stuff from behind it and what it's actually doing and creating and I'm slowly starting to build that up.
Speaker 1:But I believe, if I'm honest, the technology is moving too fast for the user, even for young people.
Speaker 2:Absolutely. No, Young
Speaker 1:people are actually frightened of AI. They don't like it. I know. So end up the young people go, I don't wanna I don't wanna be anywhere near it. Blah blah blah.
Speaker 2:Yeah. I had a a fun conversation with my daughter. She can't actually remember the time where chatty PT wasn't the thing.
Speaker 1:She's Because of her age or because she But I mean,
Speaker 2:she she must have. But, yeah, maybe she just was being silly. But
Speaker 1:Yeah. She
Speaker 2:was like, yep. Why didn't we we were talking about, I don't know, middle ages or something like, why didn't they just look it up on chat GPT? Well, I don't know.
Speaker 1:I I think I've told you this before. I've had a bit more of an awareness of five or 600 years old and a thousand year old businesses being real businesses. I think I told you about the printing place in Antwerp. So the plantain place, the whole house is still there and you read some of the letters and they were making business decisions 500 ago. Yeah.
Speaker 1:You know, selling books.
Speaker 2:Basically, the same as we do now.
Speaker 1:Yeah. Yeah. The guy printing books, it could only... Because it was all done by hand, every page, he could create maybe a 150, 200 books a year. Yeah.
Speaker 1:And there was two stories. One was this letter from his wife that was managing the business while he did the traveling. Mhmm. And she went, okay. Thank you very much.
Speaker 1:I got yours package. And the three diamonds that you got for the book you sold in Africa, I've now passed the diamonds on to your father to get them exchanged for cash. Yeah. So people were buying. I mean, that's the rarity value of it.
Speaker 1:It was just amazing. But for maps, I I don't know whether you saw my map upstairs on the wall of Antwerp from 1945. I haven't. But the ones from '16 and 1700, you had these cartographers that hand drew the map. Yeah.
Speaker 1:But he had set a business relationship where he overprinted the names. Nice. So that he was, you know, and that Yeah. Was a You know? And you're going, but this is five, six hundred years ago.
Speaker 1:This is when people didn't even have calculators. Yeah. But they were the basic business pretext. It was really interesting and I kind of seeing more and more of that now as I realize, you know, shipping around the world in all those years, was a business that they managed and the same decisions they take now.
Speaker 2:Yeah.
Speaker 1:So will AI become just a commodity? Yes. It's bound to, isn't it?
Speaker 2:Yeah. It is. Yeah.
Speaker 1:But the reality of having a robot walk through the door and actually speak to you and have a conversation is not far away.
Speaker 2:That was part of the conversation about the SpaceX thing is that the unions on Earth, they don't want any robots here. But the robots are not meant to be here. The robots are meant to be sent to Mars where they can be mining. And because of the low gravity, we can send two tons of Mars rock to the Earth every two minutes, just all mined by robots. And that just gives us all the rare minerals that we need.
Speaker 2:Yeah. And that's gonna be massively profitable and crazy, crazy.
Speaker 1:Until they need the batteries changing.
Speaker 2:Yes.
Speaker 1:I I don't mean to be negative, and so people... I'm really not negative about AI, and I'm not nervous or worried about it, and I literally do use AI every single day. Yeah. But I do wonder how organizations are going to change. All they're doing at the moment is saying, hey.
Speaker 1:I can save 200 people, so I'm gonna fire them and move them on or get them to do something else.
Speaker 2:But all the big companies that moved all those 10,000...
Speaker 1:They're realizing there's They're a of a
Speaker 2:all rehiring everyone.
Speaker 1:Yeah. Because their business case doesn't mesh out. Yeah. And it's mainly because people can't change quick enough.
Speaker 2:Yeah. It's not an AI problem. It's a people problem.
Speaker 1:A human problem. Yes. So there we go. So maybe that's what we'll talk about next time. Yeah.
Speaker 1:AI and the human problem.
Speaker 2:Yes. Alright.
Speaker 1:Cool. Oh, it's good. I enjoyed this immensely. And I mean, obviously, the idea was to talk about Seattle and where that's going. So now I'm gonna turn the microphones off.
Speaker 1:You can legitimately tell me all about the cool stuff.
Speaker 2:Yes. Let's do that.