1 00:00:00,000 --> 00:00:19,440 monkey patching podcast where we go bananas about all things tiny models coaching and more 2 00:00:19,440 --> 00:00:24,260 my name is Murillo I'm joined by my friend Bart hey Bart hey Murillo hey nice new place 3 00:00:24,260 --> 00:00:31,500 new place place so for people going places in the world yeah that's not that's not true we are going 4 00:00:31,500 --> 00:00:38,220 places we started in Murillo's closet yes we came out of the closet now we're kind of in an office 5 00:00:38,220 --> 00:00:43,180 kind of in an office yeah yeah indeed so you may hear like uh I feel like we still need to do some 6 00:00:43,180 --> 00:00:46,860 I feel like it's a bit echo you know yeah we probably need what do we need we actually have a 7 00:00:46,860 --> 00:00:52,680 we have some fur which maybe need more fur maybe more fur cool but yeah now we're gonna it's a nice 8 00:00:52,680 --> 00:00:56,400 place I'm excited about this change definitely it's gonna be good what do we have for this week 9 00:00:56,400 --> 00:01:05,300 Bart we have a lot right we do have a lot we have a lot to get to I'll kick it off we have a tiny 7 10 00:01:05,300 --> 00:01:11,400 million parameter network called the tiny recursion model claims big reasoning gains by looping on its 11 00:01:11,400 --> 00:01:20,300 own answers it reports 45 on ARK AGI 1 and 8 on ARK AGI 2 challenging bigger is better reflex with 12 00:01:20,300 --> 00:01:25,220 recursive updates if small models can reason cheaply startups get new room to build clever 13 00:01:25,220 --> 00:01:33,720 agents without big tech budgets what is this about so this is a paper from I think Samsung I want to say 14 00:01:33,720 --> 00:01:41,020 people from Samsung so we can take a look here yeah and basically they they propose a new architecture 15 00:01:41,020 --> 00:01:48,600 so it's a they call it a tiny network and it's recurrent so in the sense that the output of the 16 00:01:48,600 --> 00:01:53,800 previous iteration goes back in and it's also reasoning model right so I'm putting a lot of 17 00:01:53,800 --> 00:01:58,280 things together here so recurrent means that the output comes back in that's not how LLMs work and 18 00:01:58,280 --> 00:02:03,840 reasoning model is like when you have the the LLMs they have like a scratch pad kind of right and then 19 00:02:03,840 --> 00:02:09,060 they put some things on paper but then they just use it as intermediate steps right so it's a bit 20 00:02:09,060 --> 00:02:14,560 bold putting both things combined and they actually shown that like training a new model from scratch 21 00:02:14,560 --> 00:02:19,780 with not that much data can have very good performance on some benchmarks so I think some 22 00:02:19,780 --> 00:02:26,440 benchmarks specify here like the ARK AGI and I'll talk a bit more later but they perform on par or even 23 00:02:26,440 --> 00:02:32,960 better I think I went on the the results later down the table and I think they perform even better than DeepSeq or 24 00:02:33,840 --> 00:02:41,580 Cloud 3.7 you see so they're actually model is um was one of the best ones except except for uh Grok 25 00:02:41,580 --> 00:02:48,480 actually but considering that it was very very small models very little data is actually quite 26 00:02:48,480 --> 00:02:55,020 impressive of course for other benchmarks I would imagine they would do way worse and there was some 27 00:02:55,020 --> 00:02:59,280 information there on these are these benchmarks are geometric puzzles yeah indeed so that's what I also 28 00:02:59,280 --> 00:03:03,280 wanted to touch on but maybe just real quick as well they also mentioned hierarchical reasoning so 29 00:03:03,280 --> 00:03:08,080 this is a paper that was inspired by that okay which has also has like uh it's a similar kind of 30 00:03:08,080 --> 00:03:12,420 setup but it's a bit more complicated so this is actually a simplified version the hierarchical reasoning 31 00:03:12,420 --> 00:03:17,120 model there was like two networks and it's almost like you had two different persons and one is on the board 32 00:03:17,120 --> 00:03:23,280 doing the the more daily like the more uh hands-on tasks and then the other network would be just 33 00:03:23,280 --> 00:03:26,820 kind of saying you're in the right direction keep going like this so it was also like the 34 00:03:26,820 --> 00:03:32,320 different different way of looking at it and then there was like a feedback loop that was faster like a loop that was faster on the 35 00:03:32,720 --> 00:03:39,720 on one of those and then the the higher uh hierarchy I guess like the overlooking model also had a slower one and 36 00:03:39,720 --> 00:03:44,860 then the RKGI so that's what I also wanted to to to bring here in the discussion do you know what the RKGI is? 37 00:03:44,860 --> 00:03:55,000 RKGI yeah the benchmark right yes you know what this is who looked into it so I looked as well so the RKGI so this is RK3 RKGI is basically RK stands for 38 00:03:57,000 --> 00:04:08,020 what where is it abstract and reasoning corpus and RKGI is artificial general intelligence so it's actually from created from the same guy from the created Keras I want to say 39 00:04:08,020 --> 00:04:13,000 and he was basically arguing this little video here we'll put also on the show notes 40 00:04:13,000 --> 00:04:18,480 he basically argues that a lot of these actual benchmarks it doesn't really demonstrate intelligence 41 00:04:18,480 --> 00:04:25,480 right a lot of times if you see a lot of data if you like a monkey can copy things you know like it's not what is really intelligence 42 00:04:25,480 --> 00:04:32,480 so this benchmark was designed to to tackle more of the intelligence problem and what they do is that they have so you can 43 00:04:32,480 --> 00:04:37,480 for people following the video I'm also going over the website and there's like a play humans and build AI 44 00:04:37,480 --> 00:04:42,760 and basically the idea is that they have inputs and outputs but it doesn't tell you what the rules of the game are 45 00:04:42,760 --> 00:04:46,960 and then the model needs to figure out by itself what does it need to do to actually accomplish the task 46 00:04:46,960 --> 00:04:52,840 so I played a bit with this so this is an example right so you can play as a user and then you have the commands here 47 00:04:52,840 --> 00:04:59,400 so just uh let me share this right screen so you have like a little a little interface here you have like the the arrows so that what you can do 48 00:04:59,400 --> 00:05:04,440 and then you kind of have to like play around and then see what is actually the goal right so in the beginning you don't know 49 00:05:04,440 --> 00:05:08,980 and the idea is that you need to that you need to let the model needs to basically figure out what it is 50 00:05:08,980 --> 00:05:15,480 and that's how he defines that that's intelligence right interesting and uh apparently these for these types of things 51 00:05:15,480 --> 00:05:20,380 well that's what I'm assuming from the the paper right the the tiny network the tiny reasoning 52 00:05:20,380 --> 00:05:28,480 what's the t t t tiny networks I forgot how they called it it's like the tiny recurrent network dot t r m 53 00:05:28,480 --> 00:05:34,240 they're actually good at the tiny recursion model yeah tiny recursive model yes exactly they uh 54 00:05:34,240 --> 00:05:39,420 they're actually good at these things which again doesn't mean that it replaces a chat gpt right 55 00:05:39,420 --> 00:05:43,620 because maybe actually a lot of the stuff that we asked the gpt is not an actual intelligence 56 00:05:43,620 --> 00:05:48,740 according to the definition of the other cube but it is a different approach and I'm also wondering if 57 00:05:48,740 --> 00:05:53,300 this actually I actually saw on the paper somewhere that they said that it doesn't scale as much so 58 00:05:53,300 --> 00:05:56,180 actually having something smaller would actually give you better performance 59 00:05:56,180 --> 00:05:59,960 so I thought it was also I also wanted to bring it here because 60 00:05:59,960 --> 00:06:04,420 I don't hear as much of the the new architectures the novel things 61 00:06:04,420 --> 00:06:08,620 and I also thought it tackled a different kind of problem I also learned about the rkji so 62 00:06:08,620 --> 00:06:09,740 what do you think 63 00:06:09,740 --> 00:06:15,360 very interesting the rkji stuff makes me like the example that you just showed makes me think about this uh 64 00:06:15,360 --> 00:06:19,400 this game uh that has been running on twitch for a while uh pokemon basically 65 00:06:19,400 --> 00:06:23,520 and I think it's a I think I think it's the game boy version of pokemon that's playing 66 00:06:23,520 --> 00:06:28,420 not 100 sure and they have like uh all kinds of models trying to get to the end 67 00:06:28,420 --> 00:06:33,180 and actually like it's super inefficient there are hundreds of different things 68 00:06:33,180 --> 00:06:36,340 and the first time that it actually successfully finished the game was with 69 00:06:36,340 --> 00:06:40,900 the release of chess gpt 5 really that's very recent but the type of game is also like 70 00:06:40,900 --> 00:06:44,420 you don't like there's probably a manual somewhere out there right 71 00:06:44,420 --> 00:06:48,800 to figure out the fitness function basically how to play this game and how to how to improve on it 72 00:06:49,400 --> 00:06:53,920 but they didn't give them the the like the instructions to the models they just kind of say 73 00:06:53,920 --> 00:06:57,800 this way these are the inputs that you can do and you can monitor the output 74 00:06:57,800 --> 00:07:02,060 yeah exactly and I'll give you rewards based on the output but I won't tell you exactly 75 00:07:02,060 --> 00:07:04,480 what led you to that reward okay 76 00:07:04,480 --> 00:07:09,620 it's interesting to see these new uh architectures it's actually the first time that I'm 77 00:07:09,620 --> 00:07:17,360 I'm reading on this um yeah so that that was a the beginning when I was trying the the game right 78 00:07:17,360 --> 00:07:23,660 they said that the the the model was like easy for humans hard for ai and then I was supposed like 79 00:07:23,660 --> 00:07:28,440 this is fucking hard I was I really like no kidding I was really like reflecting like am I 80 00:07:28,440 --> 00:07:34,200 intelligent or am I like a well-trained monkey you know just repeat stuff like I had a bit of an 81 00:07:34,200 --> 00:07:40,840 existential crisis but I figured it out eventually but I do wonder how they because it's apparently good 82 00:07:40,840 --> 00:07:45,240 these type of quote-unquote puzzles yeah in games without instructions I do wonder how they train it 83 00:07:45,240 --> 00:07:52,460 on this yeah it's true whether there is no risk of over training on this like are they also testing this 84 00:07:52,460 --> 00:07:57,280 on yes I'm wondering how they how they do it's probably in the paper right probably the paper yeah 85 00:07:57,280 --> 00:08:02,660 need to need to fine comb through it but because they also different version of the benchmark so maybe 86 00:08:02,660 --> 00:08:07,460 they also trained on one before or not or I don't know if it's just actually it's a good question 87 00:08:07,460 --> 00:08:11,960 because also training in the unless you train a different version of the bank benchmark but it's 88 00:08:11,960 --> 00:08:17,840 also kind of cheating in a way I don't know yeah and I would I would also maybe because they're saying 89 00:08:17,840 --> 00:08:22,520 that these type of models score way better on these benchmarks I would maybe also argue that 90 00:08:22,520 --> 00:08:27,260 alums are typically not optimized for this type of contract yeah yeah indeed indeed and that's what 91 00:08:27,260 --> 00:08:32,360 I was also there was also I was thinking like uh I understand what the relevance of this benchmark is 92 00:08:32,360 --> 00:08:37,220 but at the same time it's like you're being very narrow about your definition of intelligence 93 00:08:37,220 --> 00:08:42,460 at the other side like there's there's uh these are seven million parameter models which are super 94 00:08:42,460 --> 00:08:47,820 efficient yeah you compare it to what is out there so that's that's definitely indeed indeed for going 95 00:08:47,820 --> 00:08:53,560 further in this field for sure even the the other model that they mentioned like the hierarchical 96 00:08:53,560 --> 00:08:59,420 something something hierarchical reasoning model the hrm is it was also like 25 million parameters 97 00:08:59,420 --> 00:09:05,780 so we're also they show they work well but this is even a step yeah smaller right so yeah again I 98 00:09:05,780 --> 00:09:11,460 think it's uh that was pretty interesting again not sure if they have a huge impact on our day-to-day 99 00:09:11,460 --> 00:09:18,260 yet but uh who knows probably some I probably would have said the same thing about transformer 100 00:09:18,260 --> 00:09:22,540 architecture when it just came out right and now we're here true all right 101 00:09:22,540 --> 00:09:30,300 else python 3.14 is out python 3.14 is free threading no gill finally lets plain threads run cpu 102 00:09:30,300 --> 00:09:36,020 work in parallel not just io lucas a friend of ours it's for model inference and async apps and shows how 103 00:09:36,020 --> 00:09:41,900 to try it out fast throughput isn't everything if latency matters this could shift python's cost 104 00:09:41,900 --> 00:09:49,380 and architecture math more than any other auto scaling button um python 3.14 is out did you know that 105 00:09:49,380 --> 00:09:55,300 i did know that it was out but i haven't really looked into what exactly was released with it yes 106 00:09:55,300 --> 00:10:00,800 so one of the things that was released well with an asterisk and i saw i hadn't seen this for myself i 107 00:10:00,800 --> 00:10:05,860 just saw like people on my network sharing about this that actually the free threading the no gill 108 00:10:05,860 --> 00:10:18,140 uh is actually out with an asterisk it's not the default yet you need to kind of set some flags and i don't know if you need to rebuild it but like maybe you don't need to build it but like you need to specify that i want the free threading version of the gill 109 00:10:18,140 --> 00:10:22,280 of the gill what is the gill maybe for people that this is the first time that they hear 110 00:10:22,280 --> 00:10:29,840 we're a bit on spot here i'm trying to say to get a eli 5 answer on this so the gill is the stands for the global interpreter lock 111 00:10:29,840 --> 00:10:38,220 and it basically means that in python even though it sometimes feels different but it can only do 112 00:10:38,220 --> 00:10:40,060 one thing at a time ever 113 00:10:40,060 --> 00:10:46,120 whenever something gets done basically locks the cpu just to that 114 00:10:46,120 --> 00:10:49,360 just to that job and then releases it and then another job 115 00:10:49,360 --> 00:10:49,860 exactly 116 00:10:49,860 --> 00:10:52,780 even if today or let's say before 3.14 117 00:10:52,780 --> 00:10:54,720 you use something like threading 118 00:10:54,720 --> 00:10:56,640 there was already a threading library 119 00:10:56,640 --> 00:10:58,920 and it felt like threading but in essence 120 00:10:58,920 --> 00:11:00,520 it was not really threading 121 00:11:00,520 --> 00:11:05,580 yeah the way i i understand is like there were different threads but like you weren't doing stuff in parallel 122 00:11:05,580 --> 00:11:07,660 it would just switch between threads all the time 123 00:11:07,660 --> 00:11:10,020 so it was a bit useless in a way 124 00:11:10,020 --> 00:11:17,100 and yeah i know python is popular with ai machine learning all these things but it's also because they 125 00:11:17,100 --> 00:11:22,040 enter they have an interface with c and then when you're on c land then you don't have this problem anymore 126 00:11:22,040 --> 00:11:31,000 that's why like numpy and all these things um got really popular right like the with python so but now 3.14 is out the 127 00:11:31,000 --> 00:11:34,680 this is a article from lucas valatka 128 00:11:34,680 --> 00:11:36,760 ex-colleague and a friend of ours 129 00:11:36,760 --> 00:11:43,800 and he kind of he makes some some interesting points on why we should care about 130 00:11:43,800 --> 00:11:51,240 about this development he also points out that now python is probably one of the fastest interpreted languages right 131 00:11:52,040 --> 00:11:54,040 big statement is it true 132 00:11:54,040 --> 00:11:56,600 i have a good question i haven't checked 133 00:11:56,600 --> 00:11:57,340 but 134 00:11:57,340 --> 00:12:00,560 on his actual arguments right so for example he says 135 00:12:00,560 --> 00:12:03,900 maybe people are pretty skeptical right so you can use multi 136 00:12:03,900 --> 00:12:06,960 i can use multi-processing which basically 137 00:12:06,960 --> 00:12:10,680 python because he couldn't run stuff in parallel 138 00:12:10,680 --> 00:12:13,600 you could run different python processes for each one of those 139 00:12:13,600 --> 00:12:16,280 like that was a way to to bypass the the gill 140 00:12:16,280 --> 00:12:20,160 right um and then he kind of says yes but there are of course 141 00:12:20,160 --> 00:12:24,080 like costs from this right can because you have different processes that are separate 142 00:12:24,080 --> 00:12:26,080 you cannot have really good communication between them 143 00:12:26,080 --> 00:12:30,080 right the other says about the the c and rust extensions um 144 00:12:30,080 --> 00:12:34,960 but then again you have to write c rust right think io so 145 00:12:34,960 --> 00:12:39,840 even though if you have something that is io bound meaning that is like you have to read 146 00:12:40,640 --> 00:12:43,040 read things in and write stuff out then 147 00:12:43,040 --> 00:12:50,560 python can be faster for these things because basically you can tell the the single thread to do 148 00:12:50,560 --> 00:12:54,880 other things while it waits for response and then it just comes back and check right but a lot of the 149 00:12:54,880 --> 00:12:59,600 stuff is actually not io bound right and then he goes on and on for like a lot of different 150 00:12:59,600 --> 00:13:05,520 articles on like okay you can scale horizontally yes but you're still very efficient on the single 151 00:13:05,520 --> 00:13:11,520 on a single instance and again machine learning inference i think a lot of this stuff also seems to 152 00:13:11,520 --> 00:13:16,880 see in rust extensions and existing io apps maybe one thing i also wanted to to to add on this 153 00:13:16,880 --> 00:13:23,680 python just because it does have the the free threading right it doesn't really mean that 154 00:13:24,880 --> 00:13:30,160 the code that you have today would just run faster no true so i do think there's a lot of code that 155 00:13:30,160 --> 00:13:36,560 would need to be reworked to be able to leverage these things i'm also wondering how many how much 156 00:13:36,560 --> 00:13:40,960 people are going to go for like how much people are going to really invest the time to to work on 157 00:13:40,960 --> 00:13:46,720 the free threading for python you know because most of the stuff that there was a real need 158 00:13:46,720 --> 00:13:52,320 people have found workarounds for this today right and also i feel like if you're going for speed 159 00:13:52,320 --> 00:13:57,920 not sure if python would be the first choice either right i think it depends a bit on the on the use 160 00:13:57,920 --> 00:14:02,480 case you're building of course i think from when you're building from scratch in the future and 161 00:14:02,480 --> 00:14:10,800 threading odds value be it be it efficiency you will probably use use this native approach right 162 00:14:10,800 --> 00:14:17,040 yeah instead of going for multi-processing i'm still curious to see i'll still be curious to see how 163 00:14:18,560 --> 00:14:23,200 how much work is it to actually implement this right because i feel like then you seem to 164 00:14:23,200 --> 00:14:27,600 manage different processes and like it's not easy right it's not like you just turn a switch and it 165 00:14:27,600 --> 00:14:32,640 just works no no you need to develop specifically for it right yeah but i feel like that's also goes 166 00:14:32,640 --> 00:14:36,480 a bit against why python is so popular that's what i'm thinking you know python is really simple you 167 00:14:36,480 --> 00:14:39,760 don't have to worry about much stuff that's why a lot of people just do it and if you do need to 168 00:14:39,760 --> 00:14:44,000 worry about all these things then you have the c extensions the rust extensions and all these things 169 00:14:44,000 --> 00:14:48,400 yeah i think it really depends on this case i think uh back in the day a long long time ago 170 00:14:48,400 --> 00:14:53,840 threading was already very popular and like even though it wasn't like need actual threading but like 171 00:14:53,840 --> 00:15:01,040 for uh for when you were building a let's say a desktop application and you had to uh you had to manage 172 00:15:01,040 --> 00:15:06,720 uh updates to the window paints and stuff like that like in the end you were using threading there 173 00:15:07,520 --> 00:15:12,640 like it's been used since day one i think what they're trying to do is simplification i think 174 00:15:12,640 --> 00:15:18,000 simplification is always a good thing if you don't need to think about anymore that reason even is a 175 00:15:18,000 --> 00:15:24,000 gill yeah there is a global interpreter lock like threading is just actual threading i think that is a 176 00:15:24,000 --> 00:15:29,840 good thing but it goes a bit i think why did it take so long is because probably for simplification 177 00:15:29,840 --> 00:15:35,920 reasons at the very early days there was a gill and that made it very hard and not very simple to 178 00:15:35,920 --> 00:15:40,320 move away because a lot of the things that were implemented like assumed that there was a global 179 00:15:40,320 --> 00:15:44,960 interpreter lock so it was not simple to move away but in the end we get to a simplification of of 180 00:15:44,960 --> 00:15:49,200 understanding the api that you're interacting with as a developer so i think that is a good thing 181 00:15:49,200 --> 00:15:53,520 i think so too but i also think that because well as i understand 182 00:15:53,520 --> 00:15:58,960 it's not like you need to learn like it's not a different programming language right like you're not 183 00:15:58,960 --> 00:16:03,520 losing a lot of the stuff everything you have today will still work and now you have this yeah so for 184 00:16:03,520 --> 00:16:08,800 sure i'm all i'm all for it but yeah again curious to see the first serious applications that we use 185 00:16:08,800 --> 00:16:14,800 this um see how the experience is but yeah let's see and the other things also so three four point 186 00:16:14,800 --> 00:16:19,680 fourteen i think that was the first the main thing or the biggest change but there are other things 187 00:16:19,680 --> 00:16:24,000 that were changed as well did you have a look at that no no i didn't so one of the things that it 188 00:16:24,000 --> 00:16:27,920 meant so yeah okay i'll just very skim through the table of contents here on this other link 189 00:16:29,600 --> 00:16:34,480 the python repo now is a bit better so there's different colors and stuff and also if you mistype 190 00:16:34,480 --> 00:16:39,040 a word and you try to run the program the interpreter will say ah did you mean this did you mean that 191 00:16:39,040 --> 00:16:45,600 yeah right so kind of like in rust they also had this now you have t strings have you heard of t 192 00:16:45,600 --> 00:16:52,320 strings template strings exactly uh i've heard of them i've never used them so now they're released so now 193 00:16:52,320 --> 00:16:57,760 now you can use it but i think the idea here is to because people use a lot of f strings right so 194 00:16:57,760 --> 00:17:05,360 yeah yeah and now i think especially with i think because a lot of lms to be honest right i think a 195 00:17:05,360 --> 00:17:09,520 lot of the you're prompting a lot of stuff for lms so basically have this different type of string 196 00:17:09,520 --> 00:17:15,280 here which is a t string you can like it doesn't evaluate to a string directly evaluates to a template 197 00:17:15,280 --> 00:17:18,800 right and then the developer can actually choose what you want to do with this what you want to 198 00:17:18,800 --> 00:17:23,840 interpolate this you can also do some uh like checks right like to to make sure there's no 199 00:17:23,840 --> 00:17:30,560 injection or some some sorts right so now it's part of so in uh on the import lib right the standard library there 200 00:17:30,560 --> 00:17:35,920 i think that's that's those are the things that really caught my attention to be honest but 201 00:17:35,920 --> 00:17:41,440 and i do i actually have heard as well that it's also the fastest python again that's good huh even 202 00:17:41,440 --> 00:17:48,160 without the the the the now it makes me actually wonder what look is the statement like how uh how 203 00:17:48,160 --> 00:17:52,960 father interpret the language is like how would this uh how does this compare to julia for example or 204 00:17:52,960 --> 00:18:03,600 sure sure i'm not sure not sure but yeah with the food or to um what's the name a bit of the the 205 00:18:03,600 --> 00:18:12,640 superset of python mojo mojo yeah but mojo is compiled i think is mojo i think so okay my bad 206 00:18:12,640 --> 00:18:20,640 i think so yeah i think so mojo is compiled okay yeah understanding mojo a python-like compiled language 207 00:18:20,640 --> 00:18:28,560 but yeah curious how these things go indeed what's next a javascript dev made over 300 000 208 00:18:28,560 --> 00:18:33,760 dollars from light gallery by dual licensing a free gpl agpl for open use paid commercial for closed 209 00:18:33,760 --> 00:18:39,360 projects and because gpl compels any bun embedding his code to open source their whole site 210 00:18:39,360 --> 00:18:44,480 most companies buy a license that is the core of the business model he pairs that with contributor 211 00:18:44,480 --> 00:18:50,480 agreements a clean major version switch and exclusive features to nudge upgrades smart sustainability or 212 00:18:50,480 --> 00:18:53,840 tollboot on open source what do you think mariella 213 00:18:53,840 --> 00:19:02,400 i don't know i don't know actually so to to make sure i understand correctly he has two licenses what 214 00:19:02,400 --> 00:19:06,960 he did uh he makes claims that he made a lot of money i'll leave that in the middle he has this uh 215 00:19:06,960 --> 00:19:11,840 this uh javascript project called light gallery this is by the way good good advertisement for that 216 00:19:11,840 --> 00:19:17,360 it's a bit of a it's it's basically like a an image and video lightbox gallery do you know like if you 217 00:19:17,360 --> 00:19:20,800 click an image on a website and you get like a gallery of images and then you can scroll through 218 00:19:20,800 --> 00:19:25,360 them yeah so that's light gallery what he does he has a he has a very much open source license 219 00:19:25,360 --> 00:19:33,360 gpl or agpl and um he also has a paid commercial license i never really thought and the paid 220 00:19:33,360 --> 00:19:39,440 commercial license is basically to make money off this which makes sense for him right like he's putting 221 00:19:39,440 --> 00:19:44,000 a lot of time and effort in this we really have this sleek thing he was trying to find a way how can i 222 00:19:44,000 --> 00:19:49,040 build a business model around this and then the gpl license is like it's open source but so a lot of 223 00:19:49,040 --> 00:19:59,040 people use default to the mit license right okay the gpl or agpl license they they are a bit more 224 00:19:59,040 --> 00:20:06,160 restrictive right they they force you as a user of that to if you make a certain change of you use it 225 00:20:06,160 --> 00:20:10,880 to be very transparent on what you use or what you do to do with it so and he says actually this difference 226 00:20:10,880 --> 00:20:17,360 between gpl v3 and agpl v3 so he says the gpl v3 is best for libraries and frameworks 227 00:20:17,360 --> 00:20:23,760 because it's triggered by distribution so if someone includes your gpl javascript 228 00:20:23,760 --> 00:20:30,800 on their public website they must must in theory open source their entire website which commercial 229 00:20:30,800 --> 00:20:36,480 companies are never going to do and agpl v3 not really relevant for his project but he says the best for 230 00:20:36,480 --> 00:20:41,920 for sas projects for the service where a code is used on a server but never distributed so the agpl is 231 00:20:41,920 --> 00:20:46,560 basically trigger when a user interacts with that software over network and if that is the case if a 232 00:20:46,560 --> 00:20:52,560 user interacts with your agpl project over that network like the company again is forced open source 233 00:20:52,560 --> 00:20:57,280 they're all which you're not going to do yeah it's very easy because it's open source projects it's very 234 00:20:57,280 --> 00:21:02,480 easy to for a web developer to use this library it's like oh look there's school looks very nice it looks 235 00:21:02,480 --> 00:21:07,520 very sleek we need this we can't use it but license only cost like i don't i don't know honestly don't 236 00:21:07,520 --> 00:21:11,600 know what this pricing is but like it's maybe it's five euro five euros a month or something right 237 00:21:11,600 --> 00:21:15,840 like it's it's a no-brainer to use a commercial license yeah so i thought it was interesting an 238 00:21:15,840 --> 00:21:21,760 interesting point of view so basically you have like uh both are more you have a paid license which 239 00:21:21,760 --> 00:21:27,840 whatever it is and then you also have a very strict open source license and then like because the mit 240 00:21:27,840 --> 00:21:32,480 is like it's a bit like you can use but you don't need to open source anything it's like right so you 241 00:21:32,480 --> 00:21:36,160 need to acknowledge i think you just need to acknowledge yeah so then you have like a very 242 00:21:36,160 --> 00:21:41,600 strict open source one and another good option and it's okay let's just go for the the small fee one 243 00:21:41,600 --> 00:21:46,400 because we want to respect the license because we have another alternative 244 00:21:47,200 --> 00:21:50,880 to the open source one and we we're not going to open source a whole website that's that's a bit the 245 00:21:50,880 --> 00:21:57,520 idea okay and he also mentions a little bit like how he handles contributions 246 00:21:57,520 --> 00:22:04,640 where you basically have like apparently there are hidden bots to do this so when you have 247 00:22:04,640 --> 00:22:10,080 users contributing to your project that has these type of licenses where there's also commercial 248 00:22:10,080 --> 00:22:15,120 part that you you can do like you can have like a contributor license agreements so contributors 249 00:22:15,120 --> 00:22:20,960 basically let you use their code but they still own it i'm going to contribute to your project so 250 00:22:20,960 --> 00:22:24,880 you can use everything but i still own it and you also have a copyright assignment agreement that's 251 00:22:24,880 --> 00:22:30,400 basically one from the moment i commit i sign it off to you it's yours okay as a repository owner 252 00:22:30,400 --> 00:22:35,280 and apparently you have fitted bots that take care of this for you that's handy that's handy yeah and 253 00:22:35,280 --> 00:22:40,160 i would definitely imagine that like if i'm using light gallery and have this fine a small tweak that 254 00:22:40,160 --> 00:22:44,480 i want to do and it's a fit but i'm not gonna care too much like i just want to get it fixed right 255 00:22:44,480 --> 00:22:49,920 yeah yeah i think they're still there's still even though like this will probably scare off the 256 00:22:49,920 --> 00:22:55,680 true open source aficionados yeah like there will still people be that are engaged enough and i've 257 00:22:55,680 --> 00:23:02,640 actually had but it's in a little bit of a different i've had similar examples but slightly executed 258 00:23:02,640 --> 00:23:08,000 differently so i used a while ago there is this library forgot the exact name uh to basically 259 00:23:08,000 --> 00:23:10,000 template word documents from 260 00:23:10,000 --> 00:23:20,160 javascript or typescript okay and you can easily generate a word document but then to it works and 261 00:23:20,160 --> 00:23:24,160 it's the best function because you have a number of alternatives but it's the best one so you go for 262 00:23:24,160 --> 00:23:31,840 that one and i say okay you can use that for free but to also let's say i think for example inserting 263 00:23:31,840 --> 00:23:36,400 images in the word document you need to plug in then you have a commercial license so i had this 264 00:23:36,400 --> 00:23:40,640 approach where it's very much open source but yeah for some parts you need to pay but i hadn't really 265 00:23:41,520 --> 00:23:48,320 seen this explicitly like the same the same thing but it's just based on how to use it whether or not 266 00:23:48,320 --> 00:23:53,040 you're under the commercial order yeah no but i think the plug because the plugins thing also feels a 267 00:23:53,040 --> 00:23:58,960 bit like the freemium model yeah true yeah it's a good comparison i think yeah but it's true i've never 268 00:23:58,960 --> 00:24:04,480 seen anything like this um i think it's an interesting venue to explore as a 269 00:24:04,480 --> 00:24:10,960 library author that is very engaged and making sure that you have something and 270 00:24:12,160 --> 00:24:15,280 it's a creative way to build a business model around exactly and i think 271 00:24:16,480 --> 00:24:22,160 ways to support open source i think it's it's always good right like finding new ways that people 272 00:24:22,160 --> 00:24:27,680 can can make a living out of these things and still like as i think it's it's a difficult discussion 273 00:24:27,680 --> 00:24:31,680 because this is good for open source but it's a bit of a slippery slope right 274 00:24:31,680 --> 00:24:40,320 why like this is still open source because it's under gpl right yeah that's what you're saying but it's 275 00:24:40,320 --> 00:24:42,880 very clearly just chosen for that so that's 276 00:24:42,880 --> 00:24:48,960 like as a company you can't really use it unless you go for the commercial license 277 00:24:48,960 --> 00:24:52,960 it's just to nudge people that yeah i know what you're saying but the advantage of course like 278 00:24:52,960 --> 00:24:58,400 it is still an actual open source license that you can use it under even though it's very it's a 279 00:24:58,400 --> 00:25:04,160 restrictive one yeah restrictive if you unless you have an open source project as well right but i feel 280 00:25:04,160 --> 00:25:09,280 like it's really like it is a very because that well if i understand what you're saying like a lot of 281 00:25:09,280 --> 00:25:15,680 of times the open source it also like open source still moves the world and i feel like 282 00:25:15,680 --> 00:25:21,520 by having this very it's almost like water and oil right like these two like the things that are 283 00:25:21,520 --> 00:25:24,240 open source with they're going to use this license and they're going to everything's going to stay 284 00:25:24,240 --> 00:25:29,760 open source and everything that's commercial there's no it won't feedback right like you won't 285 00:25:29,760 --> 00:25:34,720 you're just you're just consuming it but then like i'm also wondering like how 286 00:25:35,520 --> 00:25:41,760 would this also scale like for for if everyone had this this setup would open source be as big as 287 00:25:41,760 --> 00:25:48,320 it is they would be probably not yeah that's probably not so yeah yeah it's interesting interesting 288 00:25:48,320 --> 00:25:54,400 to think about but i do think that if this is a recipe for success i think it would be good for 289 00:25:54,400 --> 00:25:59,200 more people that are trying to also make a living out of open source to to rely on these things right 290 00:25:59,200 --> 00:26:04,480 so it's always good alternatives are always good i feel true 291 00:26:05,520 --> 00:26:12,400 all right and up next we have the netherlands takes effective control of chinese-owned chipmaker 292 00:26:12,400 --> 00:26:18,000 and xperia citing governance failings and risks to dutch and european economic security using rarely 293 00:26:18,000 --> 00:26:24,160 invoked goods availability act ministers can reverse harmful decisions wing techs shanghai listed shares 294 00:26:24,160 --> 00:26:31,760 promptly fall 10 it's a new line in europe's tech sovereignty sand does data custody now extend to 295 00:26:31,760 --> 00:26:39,280 boardroom control what do you think bart um i think it's an interesting uh news article that popped up uh i want 296 00:26:39,280 --> 00:26:50,320 to say yesterday yes october 13th yes uh that's actually today uh popped up today but um 297 00:26:52,560 --> 00:26:58,880 this is a very aggressive move by the dutch government so a little bit of background here 298 00:26:58,880 --> 00:27:02,480 wing what's the name 299 00:27:02,480 --> 00:27:09,760 wing tech is the parent company the shanghai listed chinese company 300 00:27:11,840 --> 00:27:19,280 nexperia is a subsidiary of wing tech and that is in that i think it's that quarter resides in netherlands 301 00:27:19,280 --> 00:27:27,840 was bought by wing tech i want to say in 2016 around that so not that long ago it comes originally from 302 00:27:27,840 --> 00:27:33,920 uh philips sold by philips and they uh are big in the semiconductor industry so they make things like 303 00:27:33,920 --> 00:27:39,360 transistors and diodes and stuff like that so not super fancy stuff but like things that are used throughout 304 00:27:39,360 --> 00:27:45,440 every type of electronics that you can imagine basically from household appliances to cars whatever 305 00:27:45,440 --> 00:27:52,880 they're quite big i think they employ around a thousand people um and in the netherlands or 306 00:27:52,880 --> 00:27:58,720 i think in the netherlands yeah i'm not sure on that but they're big in the netherlands and 307 00:28:01,680 --> 00:28:08,800 what now happened is that basically they the netherlands by the use of this this act they have a name 308 00:28:08,800 --> 00:28:15,600 for the act uh good availability act yes i'm a bit at loss all the details on the act but uh 309 00:28:15,600 --> 00:28:21,600 they took away control of nexperia from wing tech so that's very big right like yeah 310 00:28:21,600 --> 00:28:28,480 and take control means like so they basically uh took over the boardroom so they say that the current 311 00:28:28,480 --> 00:28:33,040 board of directors is no longer capable of making decisions 312 00:28:33,040 --> 00:28:41,440 what the next step is i'm not sure yeah to be honest and i was looking into the details on why 313 00:28:41,440 --> 00:28:45,600 this decision was made and today the details or the public details and that are still very fake 314 00:28:45,600 --> 00:28:51,360 like it's it has something to do that that the governance of nexperia 315 00:28:51,360 --> 00:28:57,000 was something was not not an order and that had to do with a 316 00:28:58,480 --> 00:29:06,640 security on technology and ownership on technology within the netherlands and the european region 317 00:29:06,640 --> 00:29:10,960 so that's very fake right yeah yeah like cyber security stuff that's the only thing they're saying 318 00:29:10,960 --> 00:29:15,760 cyber security or something like with the the knowledge on these products that they're making 319 00:29:15,760 --> 00:29:20,560 that something was happening around that but it's still very fake what is happening well why why this is 320 00:29:20,560 --> 00:29:27,120 being done i added this because it's like to me it's a very very i think it's a bit unprecedented in the 321 00:29:27,120 --> 00:29:34,560 european region i haven't heard about this at least it's um it's a very it looks like a very big 322 00:29:34,560 --> 00:29:40,160 step up in tech that looks to be linked at least to tech sovereignty yeah like i'm afraid you're gonna 323 00:29:40,160 --> 00:29:44,400 take our tech away from us even though it's clearly owned by a chinese company so it feels very 324 00:29:45,040 --> 00:29:52,080 weird like yeah it's um it's a weird situation yeah i feel like yeah i'm curious what's the next 325 00:29:52,080 --> 00:29:56,480 it's gonna be uncovered right because there's probably gonna be more and this is very there 326 00:29:56,480 --> 00:30:03,280 should be more there should be more i don't know if it's also my my bias right but like uh chinese 327 00:30:03,280 --> 00:30:08,640 government don't and then there's always a bit like it's not government owner no sorry not chinese 328 00:30:08,640 --> 00:30:13,520 government on but like um china and maybe also the the way that the title reads right dutch government 329 00:30:13,520 --> 00:30:20,640 takes control of china owned chip maker yeah it feels a bit like uh i don't know but i do hope 330 00:30:20,640 --> 00:30:25,040 that we get more details because this is to me this is not a good thing that is happening like if you're 331 00:30:25,040 --> 00:30:30,720 conducting business in a company and you're trying to abide by everyone's laws and you're doing so 332 00:30:30,720 --> 00:30:35,520 and already since well since a long time like i said like they acquired in 2016 and then suddenly 333 00:30:35,520 --> 00:30:40,160 for whatever reason that is today still very vague just saying like you're doing something wrong with 334 00:30:40,160 --> 00:30:45,600 and it's not clear which lawyer actually breaching like which is it is it compliance related like it's 335 00:30:45,600 --> 00:30:50,640 you need to have some as a when you're running account company you need to have some legal certainty 336 00:30:50,640 --> 00:30:54,880 right like the legal system that we're in now and now we're applying it like we need to be certain 337 00:30:54,880 --> 00:31:00,000 that this is we're in this for the long run yeah it shouldn't change too much but do you think that 338 00:31:01,280 --> 00:31:07,040 like they they are not also aware either what uh why they hope they are i hope they are but i hope 339 00:31:07,040 --> 00:31:12,640 that that's what i'm saying and i think it's uh it's big news but it really requires the context to 340 00:31:12,640 --> 00:31:18,000 form an opinion on this yeah no that's true that's true but i do think like at the same time if details 341 00:31:18,000 --> 00:31:23,200 do come out and you see like what the fuck like what were they doing like then then it can be a good 342 00:31:23,200 --> 00:31:30,720 thing as well right that the government still has the the the not like the power kind of but also 343 00:31:30,720 --> 00:31:34,560 like that they're they're not passive right that there is someone that is actually making sure that 344 00:31:34,560 --> 00:31:38,880 the true so it really depends on what it depends on yeah because that's what i'm like because you 345 00:31:38,880 --> 00:31:42,960 said it's not good but at first i was like but if because also this is the first time i think we 346 00:31:42,960 --> 00:31:46,480 heard this happening so i would imagine there was something very serious right so in my head i was 347 00:31:46,480 --> 00:31:52,320 thinking like oh it's good that these things are being forced right so i kind of yeah the other way 348 00:31:52,320 --> 00:31:57,680 let's assume it has a good reason right let's assume yeah indeed indeed so we'll definitely follow 349 00:31:57,680 --> 00:32:05,360 up on this what else we have simon willison coaxes claude's code interpreter into zipping its public 350 00:32:05,360 --> 00:32:09,840 follower then open sources the problems and scripts it finds they include word powerpoint excel and pdf 351 00:32:09,840 --> 00:32:16,320 skills plus a python tool that autofills forms using py pdf great for power users awkward for opsec 352 00:32:16,480 --> 00:32:20,960 how transparent should agent superpowers be to everyone poking around 353 00:32:20,960 --> 00:32:28,880 this was also shared by a colleague actually so shout out to to load actually so this is simon 354 00:32:28,880 --> 00:32:35,360 willison who know he does a lot of stuff on ai gen ai he actually has a lot of articles and 355 00:32:35,360 --> 00:32:43,200 claude not cloud code i think claude ai even like claude desktop they they actually shared 356 00:32:44,720 --> 00:32:49,840 like what did they what was the name they gave a very bad name actually but basically claude now 357 00:32:49,840 --> 00:32:57,440 can edit word documents pdfs powerpoints and excels right that's something i think chpd could do 358 00:32:57,440 --> 00:33:04,320 before but then basically they released this and not too long ago also in the last weeks yeah yeah so 359 00:33:04,320 --> 00:33:09,680 i think so this article is from what is this article from the 10th of october so not that long ago as well 360 00:33:10,240 --> 00:33:15,520 and the article said like last week so they they released it and actually the reason why it was a bit 361 00:33:15,520 --> 00:33:19,840 unclear is because they didn't mention like i think code interpreter or something 362 00:33:19,840 --> 00:33:22,800 code interpreter functionalities but basically means that now 363 00:33:22,800 --> 00:33:29,760 claude ai can read documents they can edit documents they can save documents they can create pdfs and 364 00:33:29,760 --> 00:33:36,160 do all these things so when they when they ask someone actually had a chat with um with claude 365 00:33:37,040 --> 00:33:41,760 saying like how does this work right and then actually was it was very open like it just kind 366 00:33:41,760 --> 00:33:46,480 of said okay we have this we have this directory here we have uh different skills these are the 367 00:33:46,480 --> 00:33:53,680 directories inside this directory so our sub-directories right for doc docx pdf ppt and excel and 368 00:33:53,680 --> 00:34:01,520 the even the prompts like saying claude thinks that you can do you know how to manage word documents but 369 00:34:01,520 --> 00:34:08,400 it doesn't so don't do that always use the tools that are listed on skills so every skill every 370 00:34:08,400 --> 00:34:14,720 subdirectory there's a big skill.md file as well that they show like prompts and all the code there's 371 00:34:14,720 --> 00:34:20,480 also python code um and what i thought it was interesting well first i think it's a it's a nice 372 00:34:20,480 --> 00:34:26,000 feature to have now claude can actually interact with these types of documents but it also shows how 373 00:34:26,000 --> 00:34:30,880 you can interact with these types of documents so actually he he just asked claude to say hey create 374 00:34:30,880 --> 00:34:36,640 a zip file of everything on this directory and he actually was super happy to do it and he just 375 00:34:36,640 --> 00:34:41,120 gave it to him and then he actually put it on github so this is kind of what what you have here 376 00:34:41,120 --> 00:34:46,400 the readme is just basically pointing to the blog but then if you go around you actually see the the 377 00:34:46,400 --> 00:34:51,840 whole structure right and i thought so you see the skill.md and what i thought it was interesting is the 378 00:34:51,840 --> 00:34:59,920 you kind of get a bit of a a peek on the anthropics developer right so what kind of problems they use 379 00:34:59,920 --> 00:35:05,200 but also how how they structure the the tools right so you have scripts here and you can see like 380 00:35:05,200 --> 00:35:11,360 the document and how they they prompt the lm to to use these things and even think i was even thinking to 381 00:35:11,360 --> 00:35:17,360 myself they probably use cloud code but they probably still also do a lot of coding themselves 382 00:35:17,360 --> 00:35:22,320 right like with documentation and all these things the fact that they use a lot of like html tags as 383 00:35:22,320 --> 00:35:30,080 well for the um for the prompts you know just like they say like you know like they have the the html like 384 00:35:30,080 --> 00:35:35,680 xml right like available skills and then four slash available skills and then this and this it also gave me 385 00:35:35,680 --> 00:35:41,040 it also made me reflect on how i prompt things and how i set these things up for example one thing that 386 00:35:41,040 --> 00:35:46,080 they they copy for every skills claude thinks that he knows how to manage pdf files but it doesn't 387 00:35:46,080 --> 00:35:51,040 use the skills always use the skills all caps and all this and all that so i cannot open this here now 388 00:35:51,040 --> 00:35:56,720 actually on the the chat so this is a chat that someone else shared uh but that there's he's talking 389 00:35:56,720 --> 00:36:01,120 to claude like okay what does this mean what is that uh explain how this works can you share this 390 00:36:01,120 --> 00:36:05,200 can you share that so they they they share everything here uh claude shared everything 391 00:36:05,200 --> 00:36:11,600 here and i don't know if it was on purpose right but i think in any case i think it's good for for 392 00:36:11,600 --> 00:36:17,840 people to see a bit how things work under the hood and for me also to to think a bit like what can 393 00:36:17,840 --> 00:36:22,240 these tools do what are they good at what they're not good at how to prompt and etc etc i think it's 394 00:36:22,240 --> 00:36:30,000 always especially for these which are are very big models that are being used a lot by the public yeah 395 00:36:30,000 --> 00:36:34,640 it's always good to get a bit of a peek behind the curtain right yeah i feel like also these people 396 00:36:34,640 --> 00:36:38,880 they're working constantly with these things they're constantly building tools but i think also having 397 00:36:38,880 --> 00:36:44,400 some some like tips right from them like this is how this works how that works this is what worked 398 00:36:44,400 --> 00:36:48,560 well this is what didn't work well i thought it was very interesting one thing that also on the 399 00:36:48,560 --> 00:36:56,320 article simon willis says that they also have a lot of nice tools so like if you want to manipulate pdfs 400 00:36:56,320 --> 00:37:01,120 with python you can also just take one of the tools that they have there and just just use it right so 401 00:37:01,120 --> 00:37:06,800 there's a lot of interesting stuff there so cool i invite everyone to to have a look right 402 00:37:06,800 --> 00:37:15,680 all right what and what's next london stock exchange group says customers can now 403 00:37:15,680 --> 00:37:21,680 build copilot studio agents that use licensed data from workspace and financial analytics to an lseg 404 00:37:21,680 --> 00:37:27,760 managed mcp server it is fewer glue scripts more govern access agents live inside microsoft 365 but 405 00:37:27,760 --> 00:37:35,920 talk to lseg via model context protocol so i thought this uh for um the simple fact that we discussed 406 00:37:37,280 --> 00:37:45,040 mcp service last week yeah with the snowflake snowflake we're using the managed mcp for server this week 407 00:37:45,040 --> 00:37:50,720 it's the london stock exchange group we have a managed mcp server now through which you can uh 408 00:37:50,720 --> 00:37:57,120 basically get information on financial markets that's how i understand it um via the model context 409 00:37:57,120 --> 00:38:04,480 protocol is communicated and it is very neatly apparently in uh copilot 365 but i assume that you can also use it 410 00:38:04,480 --> 00:38:11,120 with other uh with other uh mcp uh clients the mcp server like you have like it says here microsoft 411 00:38:11,120 --> 00:38:15,840 365 but i guess you can have any client that you use the mcp server that's the idea i think in theory 412 00:38:15,840 --> 00:38:19,920 yes and in practice i don't know they probably integrated very heavily with uh with microsoft 413 00:38:19,920 --> 00:38:24,800 copilot for authentication stuff like that i would assume that's because they're their release have 414 00:38:24,800 --> 00:38:29,440 focuses heavily on microsoft yeah that's what i was also like a bit curious because mcp should be 415 00:38:29,440 --> 00:38:33,760 client agnostic but they are mentioning a lot of microsoft stuff but maybe it is for the authentication in 416 00:38:35,280 --> 00:38:39,520 yeah cool so you know what kind of what kind of information you can pull from it or no no i don't 417 00:38:39,520 --> 00:38:44,320 know it's financial market information but i've really gone in depth but i think it's yet another 418 00:38:44,320 --> 00:38:49,280 sign like we see a lot of these things like official mcp servers coming out will we have a mcp server for 419 00:38:49,280 --> 00:38:57,760 the monkey patching podcast what would the tools be that like you find find find articles find the 420 00:38:57,760 --> 00:39:02,560 quotes or finding for or maybe just asking questions about like what's the give me a summary of the 421 00:39:02,560 --> 00:39:08,160 last episode i think there's a good mcp server of just like no one really thought about like they 422 00:39:08,160 --> 00:39:13,360 they just thought about can we do it they build it and no they didn't really realize how should we do 423 00:39:13,360 --> 00:39:18,800 it it was like there's no mic but it's there but it's there manage and all that's secure manage and 424 00:39:18,800 --> 00:39:22,880 all like it's just like the message ends up in your slack and you type back exactly just like 425 00:39:22,880 --> 00:39:28,560 like automation zafir's thing murillo in the loop but no but i think i do mention last time that we see 426 00:39:28,560 --> 00:39:32,720 like it's probably going to be something that will continue happening right like i think it makes 427 00:39:32,720 --> 00:39:38,320 sense that like these providers they have their own yeah and because you also have these like we also 428 00:39:38,320 --> 00:39:44,080 discussed last time like with uh open ai's agent kit yeah the canvas builder that they have now like 429 00:39:44,080 --> 00:39:48,720 they have full support for using tools from ncp service so you know no longer only see it in 430 00:39:50,720 --> 00:39:56,160 clients like cloud like gpt like these same things or or copilot but also in workflow automation tools 431 00:39:56,160 --> 00:40:01,200 like yeah the there's really an uptake of the under of this technology now for sure for sure and i 432 00:40:01,200 --> 00:40:05,600 think as soon as you have like i think about that like i said agent kit as soon as you have like 433 00:40:05,600 --> 00:40:13,360 these low code tools or no code tools the market for like the need for these mcp service increases a lot 434 00:40:13,360 --> 00:40:17,840 yeah definitely right and i think it's good as well i think the protocol will also mature and 435 00:40:18,480 --> 00:40:23,840 other things as well so it's really really cool what else we have um a women's health editor 436 00:40:23,840 --> 00:40:28,800 trades six weeks with runa stravas new coaching app and likes the tail of paces behind this entirely 437 00:40:28,800 --> 00:40:34,160 interface amid tiktok critiques of aggressive mileage runner replies we don't use ai to generate 438 00:40:34,160 --> 00:40:40,080 training plans seeing expert designs them while ai adjusts the progress fair deal for experienced 439 00:40:40,080 --> 00:40:44,800 runners may be risky for beginners so how much coaching can an app do before your 440 00:40:44,800 --> 00:40:47,840 knees file a complaint you're a runner no bart 441 00:40:49,280 --> 00:40:52,640 i would normally say yes but maybe i could say now i used to be a runner because we've been 442 00:40:52,640 --> 00:40:57,280 engine for a shitty long time but let's not go into that but you're using runner 443 00:40:57,280 --> 00:41:04,000 that would be a good excuse yeah it would be easy if you can just blame like it like an app 444 00:41:04,000 --> 00:41:12,640 yeah it's this fucking app yeah it's like yeah it would be easy yeah now why i put this here i think 445 00:41:12,640 --> 00:41:19,280 it's interesting like it goes into bits like there's some discussion in that community that they think 446 00:41:19,280 --> 00:41:23,680 that ai is being used to generate training plans and because they typically 447 00:41:23,680 --> 00:41:31,200 these training bags go a bit too high in typical training load like a bit too intensive and that's 448 00:41:31,200 --> 00:41:37,440 apparently also what you get when you use something like chpd directly so it's very apparently very 449 00:41:37,440 --> 00:41:41,360 very but the issue is not that it's using ai is using ai without any 450 00:41:41,360 --> 00:41:46,240 layer on top of it they're just using the plain ai because i guess you could also have no i think 451 00:41:46,240 --> 00:41:52,560 their issue is that they're using ai without enough actual knowledge on yeah exactly on the training 452 00:41:52,560 --> 00:41:57,440 a theory behind it yeah yeah because that's that that's also what i was fishing for because 453 00:41:57,440 --> 00:42:03,440 you could also have you could also use ai but you can have like a curated sources yeah that you should 454 00:42:03,440 --> 00:42:07,840 instruct the ai to always follow through but that's that's that's not what apparently they're using 455 00:42:07,840 --> 00:42:11,840 well well not they're a bit vague on this i think that's a bit the the feedback from the community 456 00:42:11,840 --> 00:42:16,880 is a bit that they're questioning that like how much is actually um a app is this runner which 457 00:42:16,880 --> 00:42:22,080 was recently acquired acquired by strava not too long ago let's say four ish months ago okay so 458 00:42:22,080 --> 00:42:27,120 was it there was a startup or something and then it was a startup um i think a uk startup 459 00:42:27,440 --> 00:42:32,400 which grew very very quickly i also think they did a shit ton of marketing like a you couldn't 460 00:42:32,400 --> 00:42:37,520 look at anything sports related without getting an ad oh really wow i think they grew also a lot through that 461 00:42:37,520 --> 00:42:43,280 but we we see a lot of these apps coming up right like we also have something in belgium 462 00:42:43,280 --> 00:42:47,360 which which i would say is a competitor to runa it's called renara but you have a lot 463 00:42:47,360 --> 00:42:51,920 of these candidates right every other training tool is now building an ai coach yeah but i'm 464 00:42:51,920 --> 00:42:56,720 also wondering if it's because it's quote-unquote it's easy because i feel like if run is really 465 00:42:56,720 --> 00:43:01,680 just using chgpt then yeah like you can have a hundred of these competitors right 466 00:43:01,680 --> 00:43:08,160 just marketing around and maybe the ui or something right that's true so yeah maybe 467 00:43:08,160 --> 00:43:14,560 one thing before we continue this is on the woman's health mag is there anything woman related to this 468 00:43:14,560 --> 00:43:19,680 because i think it's just a woman that's tried runner okay it's the only thing reason why it wasn't the 469 00:43:20,560 --> 00:43:27,600 the women's health uh okay um what do you think of this uh coaching 470 00:43:27,600 --> 00:43:33,760 i think it makes a lot of sense when you do it in a very curated manner 471 00:43:33,760 --> 00:43:39,440 i think what is the the challenge if you just if i just asked just give me a play in training 472 00:43:39,440 --> 00:43:44,080 plan because in three weeks i need to do that and help me prepare for that in three weeks like it's 473 00:43:44,080 --> 00:43:47,840 not sure if that's the best support you can get then you could probably better go to 474 00:43:47,840 --> 00:43:52,880 or whether stream or three weeks or three months doesn't really matter but like without specific 475 00:43:52,880 --> 00:43:57,760 instructions it's probably better to go to coach to avoid through like a coach i mean like a person a 476 00:43:57,760 --> 00:44:04,320 person there's some actual experience to avoid things like overtraining and injuries but i think from the 477 00:44:04,320 --> 00:44:11,520 moment that you have this a theoretical knowledge on training physiology and you make sure you have a 478 00:44:11,520 --> 00:44:17,040 very curious prompt or or whatever solution to to inject that i think then it starts to become 479 00:44:17,040 --> 00:44:27,440 better i think when it makes it even better is when you have objective feedback from users or from 480 00:44:27,760 --> 00:44:31,920 from users or from wearables from the user ah yeah i see what you're saying uh because you had a 481 00:44:31,920 --> 00:44:39,120 little toy project around this no yeah it's already a while ago for uh to prepare for a race that i had 482 00:44:39,120 --> 00:44:44,240 in like uh in in two months or something yeah and i think i remember if i remember correctly correctly 483 00:44:44,240 --> 00:44:48,880 if i'm wrong so you had the ordering so which gives you like a sleep score and readiness score 484 00:44:48,880 --> 00:44:53,600 yeah and you also prompted to say i have this i want to have like i think you said like i want 485 00:44:53,600 --> 00:44:58,720 to have this much rest and this and this and this one's long dates short day and then you could give 486 00:44:58,720 --> 00:45:04,800 feedback after every after every exercise that would basically add to the same conversation and then 487 00:45:04,800 --> 00:45:10,480 you could also add like little reminders like okay i i will do the long run today and i forgot my energy 488 00:45:10,480 --> 00:45:15,600 hell whatever then like yeah right um but then you also prompted to say this is what i want this is 489 00:45:15,600 --> 00:45:20,560 kind of the same thing you didn't just say i want i have a race in three months get me ready no it was 490 00:45:20,560 --> 00:45:25,360 according to a certain uh let's say a certain framework yeah and you knew already this that's 491 00:45:25,360 --> 00:45:31,200 why i already put it in yeah so and but i think that's like i i use for example an aura ring but you 492 00:45:31,200 --> 00:45:37,440 can also use let's say your your uh your apple watch or a whoops wrap but i think it's very important 493 00:45:37,440 --> 00:45:42,080 to do this because i think as a person you're you're not very objective towards yourself 494 00:45:44,560 --> 00:45:49,200 like if you're preparing for a race and you feel like i'm not so i'm a bit tired but like it's only 495 00:45:49,200 --> 00:45:54,080 three days left to go i mean i'm just gonna do the hard worker today and and well for me it's the 496 00:45:54,080 --> 00:45:59,680 opposite it's like i'm a bit tired i'm just gonna stay yeah like to each their own right i just left 12 497 00:45:59,680 --> 00:46:03,600 hours i feel like it's like something like an o-ring or a whoop or an apple watch will probably give 498 00:46:03,600 --> 00:46:09,040 much more objective information i think for sure you need this combination because otherwise an ai can 499 00:46:09,040 --> 00:46:16,240 distinguish which personal coach would probably do like a good one yeah but i think i agree because i 500 00:46:16,240 --> 00:46:22,080 think also you have more touch points like the ai has no the only interaction the ai has for you 501 00:46:22,080 --> 00:46:26,880 is really just the type of text that you write and i think if you have a user there's tone of voice 502 00:46:26,880 --> 00:46:31,760 there's body language there's there's a lot of more stuff and i feel like if you you just a person 503 00:46:31,760 --> 00:46:36,480 has more context than just the ai has and i think by adding these other things you're also giving more 504 00:46:37,280 --> 00:46:41,760 context for the model right uh yeah i agree yeah and of course a lot of these things are often baked 505 00:46:41,760 --> 00:46:45,840 in when you use an app right like when you use runner the layer after training it probably asks you 506 00:46:45,840 --> 00:46:51,280 like how it went and stuff like that it's still your input it's not as objective as yeah that's your 507 00:46:51,280 --> 00:46:57,600 recovery matrix yeah and uh maybe but i do believe that because i'm saying like it's probably better to 508 00:46:57,600 --> 00:47:03,360 go to a coach i'm saying in this situation it's probably better to go to experience coach for sure i think 509 00:47:03,360 --> 00:47:09,680 your below average coach it's probably better to use ai right yeah yeah so i'm not i'm not a runner 510 00:47:09,680 --> 00:47:14,640 right i did i did two sports the more tennis and football for example but i'm not a runner and i feel 511 00:47:14,640 --> 00:47:20,960 like it's very different disciplines right but maybe to be the devil's advocate here like is it that 512 00:47:20,960 --> 00:47:27,680 different like so you could also argue that the training data from the llm right some of it was also 513 00:47:27,680 --> 00:47:33,840 diet was also running routines like you know like and i would imagine that it's not like every three 514 00:47:33,840 --> 00:47:37,760 months there's a new running regimen it's like oh yeah this is the state of the art and everything 515 00:47:37,760 --> 00:47:44,080 actually everything before was bad true so in terms of like knowledge is that just the training data of 516 00:47:44,080 --> 00:47:50,400 the lm not enough compared to like a let's say an average coach would the average coach be able to 517 00:47:50,400 --> 00:47:54,800 give you better information no with the same i think an average coach 518 00:47:54,800 --> 00:48:01,680 better understands how you're feeling so you would just be able to ask more of the questions 519 00:48:01,680 --> 00:48:05,920 and like to to get the right content so it's not about like if you have the same context for both 520 00:48:05,920 --> 00:48:10,560 they do probably a similar job but the coach you're talking about an average coach now every 521 00:48:10,560 --> 00:48:16,400 coach yeah but the even the average coach she would be able to better poke for the right context 522 00:48:16,400 --> 00:48:20,320 like how you're feeling okay you're feeling a bit tired but like is it tired because you didn't sleep 523 00:48:20,320 --> 00:48:24,320 well or is it because you have do you have pain but how is the pain and maybe an lm wouldn't do that 524 00:48:24,320 --> 00:48:30,160 as much that's the and i think also especially when face-to-face you have a lot of non-verbal 525 00:48:30,160 --> 00:48:34,400 communication yeah maybe you have a limp or maybe you're just a bit like you look maybe it's a bit 526 00:48:34,400 --> 00:48:39,600 too much but dragging yourself and you're walking in a wheelchair fine you want a wheelchair you know 527 00:48:39,600 --> 00:48:43,760 like hello um but i get what you're saying i get what you're saying also like maybe you can also 528 00:48:43,760 --> 00:48:49,440 perform some assessments like okay does it hurt like try to like push here resist my push like 529 00:48:49,440 --> 00:48:53,360 does this hurt does that hurt etc etc or also there are some things i also think it's very hard to 530 00:48:53,360 --> 00:49:00,320 explain to ai right like i it it my knee hurts when i turn it like this you know like how do you like to 531 00:49:00,320 --> 00:49:03,840 you know you have someone that's like there you're just showing and it's like yeah that's what you're 532 00:49:03,840 --> 00:49:09,120 pointing like right here right here when i turn from here to here and i'm looking like towards the 533 00:49:09,120 --> 00:49:14,800 sun it kind of like yeah um so i think that's also useful i think it's also very very useful 534 00:49:14,800 --> 00:49:24,400 do you use any uh coaching uh running app or anything no i have an actual uh human coach did you ever 535 00:49:24,400 --> 00:49:29,600 try the actual uh aside from your little toy project 536 00:49:29,600 --> 00:49:34,080 no only the toy projects 537 00:49:34,080 --> 00:49:39,680 do you think there is a future for this do you think like maybe 538 00:49:39,680 --> 00:49:44,960 no there's definitely i think there's already like it's already there's a personal base yeah but 539 00:49:44,960 --> 00:49:50,800 that's a bit of the the i think the that feels a bit like the challenge there is that the human 540 00:49:50,800 --> 00:49:56,080 coach is way more expensive i mean you easily uh in the given range here like probably around 541 00:49:56,080 --> 00:50:04,560 75 to you could go probably up to 200 euros a month yeah for a coach that follows you up yeah yeah 542 00:50:04,560 --> 00:50:10,240 personally versus paying uh what is it maybe i don't know 15 20 euros yeah for something like 543 00:50:10,240 --> 00:50:15,200 it yeah sure it's a big difference right like i think a lot of people that are not super serious 544 00:50:15,200 --> 00:50:19,920 when it comes to sports or just starting out will very quickly up for let's try this out and i'm also 545 00:50:19,920 --> 00:50:23,840 wondering like the people that are not really into running or people that are starting who do you 546 00:50:23,840 --> 00:50:28,800 think they're the ones that actually should be the more careful you know what i'm saying like maybe 547 00:50:28,800 --> 00:50:35,440 like you don't know like you're an expert runner or you were but uh like you have a lot of experience 548 00:50:35,440 --> 00:50:40,320 so maybe for you to use these apps it will be not as dangerous as someone that is just starting out 549 00:50:40,320 --> 00:50:43,920 but the person is starting out doesn't want to invest yeah right so it's a bit of a how do you 550 00:50:43,920 --> 00:50:51,360 find the right market for the right things um and maybe one last thing before i move on what about diet would you 551 00:50:53,520 --> 00:50:56,800 because when you're thinking of this i think of sports preparation and i think of like dieting and 552 00:50:56,800 --> 00:51:01,600 all these things and to me i feel like there's a lot of parallels right like dieting as well i don't 553 00:51:01,600 --> 00:51:07,680 think diet changes every month but i also feel like there's a bit of non-verbal also like trying 554 00:51:07,680 --> 00:51:12,080 these things and and having someone talking to you and asking the right questions do you think they're 555 00:51:12,080 --> 00:51:21,120 different like if the question is is llms or ai a good use for coaching running we discussed but like for 556 00:51:21,120 --> 00:51:26,640 diets or giving advice for a dietitian things about the same things different i think the objective 557 00:51:26,640 --> 00:51:31,920 measurements are maybe a bit easier like the result like you have the weight yeah but i feel like you 558 00:51:31,920 --> 00:51:35,920 can also lose weight and healthy right i agree i agree but like i think that it's the same challenge 559 00:51:35,920 --> 00:51:41,040 because whether you're you're coaching a performance or you're coaching diets you still need 560 00:51:41,040 --> 00:51:47,840 something that coaches you and i think if we're actually talking about coaching i hope i believe 561 00:51:47,840 --> 00:51:53,200 that you myself still better at coaching than in the ice today yeah yeah i think today for sure and i think 562 00:51:53,200 --> 00:51:59,760 any i can support but then you need to like do part of that coaching on yourself basically yeah 563 00:51:59,760 --> 00:52:05,920 but i think and i think we talked some weeks ago about the hallucination thing of open ai 564 00:52:05,920 --> 00:52:13,200 i think the biggest biggest biggest the easiest reason that i can give to anyone why a human will always be 565 00:52:13,200 --> 00:52:18,560 better or not always but our place better today is that llms they don't ask as much follow-up questions 566 00:52:18,560 --> 00:52:22,640 they don't know when to ask questions and when to ask and i think for a human they always they will 567 00:52:22,640 --> 00:52:26,960 have them more like what do you mean by this what do you mean by that you're saying this but maybe 568 00:52:26,960 --> 00:52:30,800 what you mean is that like do you have pain or do you have soreness right like all these things 569 00:52:30,800 --> 00:52:38,720 and uh i think lms will definitely fall short on these things all right moving on we have post hog 570 00:52:38,720 --> 00:52:44,720 lays out six easy to make ai coding mistakes think big code code based blindness in context and quote 571 00:52:44,720 --> 00:52:56,320 unquote led the agent to everything optimism their code base spends 8,984 files and 1,623,533 lines 572 00:52:57,280 --> 00:53:04,560 and one engineer jokes and i quote cloud code riding rust is a while loop that accelerates climate change 573 00:53:04,560 --> 00:53:11,760 so lock in guardrails cursor rules spec files linkedin and expect more code more reviews and yes 574 00:53:11,760 --> 00:53:22,080 more bugs if you don't curious about this one bart you share this uh yeah this is from the post hoc team yes 575 00:53:25,040 --> 00:53:32,640 they are uh they uh basically made uh an overview on um how to avoid common mistakes when 576 00:53:32,640 --> 00:53:39,920 when using tools like cloud code or cursor i'm a big fan of postdoc i think it's cool it's cool that 577 00:53:39,920 --> 00:53:44,000 they are publishing stuff like this so postdoc for people that don't know it is like this uh 578 00:53:44,000 --> 00:53:50,560 quote-unquote a bit of an uh all-in platform if you're building a product or a web app because you can 579 00:53:50,560 --> 00:53:55,200 get if you use postdoc it's very easy to plug in and then you get a lot of analytics on how your product 580 00:53:55,200 --> 00:54:00,720 is being used uh like how often is this button clicked or a specific user segments that use specific 581 00:54:00,720 --> 00:54:07,600 functionality more and more often um you can do easily do a b testing you can have feature flags so 582 00:54:07,600 --> 00:54:13,600 let's say for this user segment we show this feature for that user segment not or so it's and it's super 583 00:54:13,600 --> 00:54:20,080 super intuitive super easy the website is really cool the website they're a bit of uh tongue-in-cheek 584 00:54:20,080 --> 00:54:25,280 yeah uh a bit very very much developer focus yeah makes me think a bit about like model duck 585 00:54:25,280 --> 00:54:28,480 a little bit same type of uh tongue-in-cheek humor 586 00:54:31,200 --> 00:54:37,200 recently yeah um i think i want to say a week ago released this blog post and they say 587 00:54:37,200 --> 00:54:45,600 we'll quickly go over the remarks that they make you should not treat your big code base like a small 588 00:54:45,600 --> 00:54:52,320 code base which i guess i guess good feeling makes sense but i think it's if you work on a small code 589 00:54:52,320 --> 00:54:58,480 base like a few files but you start up cloud code or codex or whatever and you just ask it to make changes 590 00:54:58,480 --> 00:55:05,600 and it typically works and you are i think typically it's it's a danger to just switch to the big 591 00:55:05,600 --> 00:55:11,040 project to the same yeah i would i haven't done that yet but i i can imagine because i think also yeah 592 00:55:11,040 --> 00:55:16,960 context and size and like not duplicating things that already exist yada yada right yeah and i think they're 593 00:55:16,960 --> 00:55:24,160 like the then comes to the second thing like the you need to from the moment that you end up in 594 00:55:24,160 --> 00:55:28,320 these big costs you need to provide right context like specific rules specific guardrails it depends a 595 00:55:28,320 --> 00:55:34,960 bit on what type of uh that's all you're using how you set this up right like you use cloud on d or 596 00:55:34,960 --> 00:55:40,080 yeah or cursor rules stuff like this third thing is trying to use the i have something you know is not 597 00:55:40,080 --> 00:55:46,880 good at so they they say here that it's for example uh not good at rust um or specific specific 598 00:55:46,880 --> 00:55:52,720 languages or specific niche dialects is there for example the hawk query language like don't use it for 599 00:55:52,720 --> 00:55:59,600 stuff it typically hasn't seen too much um makes sense but then also like how do you know what what 600 00:55:59,600 --> 00:56:03,680 it's good and what's not good just try it out i think it's a bit uh experimentation yeah it's also 601 00:56:03,680 --> 00:56:07,040 something that you probably should reevaluate as time passes yeah yeah for sure 602 00:56:07,040 --> 00:56:14,000 being content with your existing workflow i think we uh all very quickly fall into that gap like this 603 00:56:14,000 --> 00:56:18,400 works but maybe we should try this other thing and then actually have like a degradation of the 604 00:56:18,400 --> 00:56:21,520 performance right yeah um i have a 605 00:56:22,720 --> 00:56:26,960 i have i have that problem for sure like oh maybe i should try this i'll just try that i think for 606 00:56:26,960 --> 00:56:31,920 editing the podcast as well something that i definitely like i have like a a setup and it's 607 00:56:31,920 --> 00:56:36,480 like ah but maybe i can do this i can use this feature i can do this and it's like no just just 608 00:56:36,480 --> 00:56:43,200 just just just do it yeah exactly um another thing you mentioned here is not using ai it's not a good 609 00:56:43,200 --> 00:56:50,320 idea okay um even if you dislike it personally you should still realize like your competitors are using 610 00:56:50,320 --> 00:56:56,960 ai i see so the so not using i is the mistake is that you should use yeah yeah exactly okay um 611 00:56:56,960 --> 00:57:01,440 and your users are almost certainly using it maybe the exception here or there true 612 00:57:01,440 --> 00:57:06,960 which is a fair point but then like is the argument here saying like 613 00:57:06,960 --> 00:57:12,720 because i makes you more productive and your competitors are user and your users are using or like 614 00:57:12,720 --> 00:57:19,120 why is this i think your competitors are using it your users are using it like you need to know what 615 00:57:19,120 --> 00:57:24,880 is going on in the world and to i see yes but it's very specific to post hog right because they are 616 00:57:24,880 --> 00:57:31,120 developer focused like they don't have a product for marketing when it comes to users maybe but even 617 00:57:31,120 --> 00:57:38,080 there like like if you have like this i don't know marketing product that makes uh but that's whatever 618 00:57:38,080 --> 00:57:45,040 generates uh social posts for you like competitors are also using it right yeah your users are probably 619 00:57:45,040 --> 00:57:48,640 expecting that they can type something and like something gets templated up by an ai 620 00:57:48,640 --> 00:57:54,640 right like it's they think you can make this he's more than a lot of different industries so i think 621 00:57:54,640 --> 00:57:58,720 what they're saying is like you should always be touching ai because you need to know what it can do 622 00:57:58,720 --> 00:58:02,480 because people are going to expect you to be able to things that you can do so just being like 623 00:58:02,480 --> 00:58:07,280 by doing it you're always going to be up to date it's a good it's a good point i hadn't thought of 624 00:58:07,280 --> 00:58:12,480 that and another bad idea is not is uh is letting ai do everything for you it's the other the other 625 00:58:12,480 --> 00:58:18,080 opposite right zero extreme yeah yeah which which i guess makes sense i think that it's uh 626 00:58:18,080 --> 00:58:23,520 sometimes a bit the challenge that uh because you use it a lot you also quickly try to use it in 627 00:58:23,520 --> 00:58:27,680 different situations even though you know it's not always the best but just to be aware of that 628 00:58:27,680 --> 00:58:32,080 so i think it's an interesting one we'll post the link of the article and show notes anything that 629 00:58:32,080 --> 00:58:37,120 surprises you on this not necessarily um i think you're good feeling wise i think i think the one 630 00:58:37,120 --> 00:58:42,800 about not using ai is not a good idea i think that is becoming the reality i think you still have 631 00:58:42,800 --> 00:58:47,760 people with a very strong opinion on not wanting anything to do with ai that i have the feeling that 632 00:58:47,760 --> 00:58:56,080 it makes yourself irrelevant but i think even like i like to we had all mit study that 95 of stuff fails 633 00:58:56,640 --> 00:59:01,200 i also saw again the other i think it was mit study as well that like using chat gpt will make your brain 634 00:59:01,200 --> 00:59:11,120 whatever yeah but i think like the reality is that today ai is getting ingrained in whatever we do 635 00:59:11,120 --> 00:59:15,760 yeah whether for good or for bad i don't know like you can argue about it look we had the example of 636 00:59:15,760 --> 00:59:21,680 this running a running app right yeah like it used to be like it used to be just programmatic like and 637 00:59:21,680 --> 00:59:26,640 now every running apple probably have an option to have an ai generated plan and i think often you 638 00:59:26,640 --> 00:59:34,960 you don't even see anymore the ai gets used behind the that is simply the reality right it's getting 639 00:59:34,960 --> 00:59:39,520 embedded in everything for better or for worse and i think probably when we look back in 10 years a 640 00:59:39,520 --> 00:59:44,000 lot of these products have become a lot better than they are now for sure they will just keep getting 641 00:59:44,000 --> 00:59:49,600 better i mean the people are using as well yeah and also the the features ai enables yeah yeah 642 00:59:49,600 --> 00:59:54,000 through that definitely and i think this is a healthy stance right like don't do everything 643 00:59:54,000 --> 01:00:00,000 don't try to do everything but do something watch yourself yeah the last thing is you said like 644 01:00:00,000 --> 01:00:06,240 challenging yourself as well i also i heard it somewhere and i also think it's true broader than just 645 01:00:06,800 --> 01:00:15,200 programming right or whatever that i think it's healthy to not be married to your beliefs right like 646 01:00:15,200 --> 01:00:22,080 always be like maybe i'm wrong maybe let's try this maybe let's try that i think it's just healthy for 647 01:00:22,080 --> 01:00:27,120 you as a person you know to have that attitude towards life you know like yeah i agree like i think if 648 01:00:27,120 --> 01:00:31,760 you ever hear yourself saying i don't agree and i don't care what you say there's nothing you can 649 01:00:31,760 --> 01:00:35,920 say to convince me that's not healthy right there should always be something that i can say to you 650 01:00:35,920 --> 01:00:40,080 know yeah i fully agree yeah and it works the other way around as well and i think we shouldn't 651 01:00:40,080 --> 01:00:44,560 every now and then have a look at like we're discussing a lot like where is this actually helping 652 01:00:44,560 --> 01:00:49,440 about true like yeah and i think that's the other the other extreme quote-unquote right like uh 653 01:00:49,440 --> 01:00:58,960 maybe i is not the way to go and i'm happy to to consider that possibility as well right so let's see well let's see 654 01:01:01,760 --> 01:01:08,960 that is it for the articles we have today yeah um we have a little bit of news as well we have a 655 01:01:08,960 --> 01:01:14,400 sign-up link we have a sign-up link or tell me more about the sign you can go to newsletter dot 656 01:01:14,400 --> 01:01:21,040 monkeypatching.io and you can basically leave your email address what are you gonna do with this email 657 01:01:21,040 --> 01:01:25,040 you're gonna send well that's a big question you're gonna sell this to people to yeah we're 658 01:01:25,040 --> 01:01:30,720 gonna sell their data i'm gonna get rich off it yes um first thing we'll buy is a yacht yeah okay 659 01:01:30,720 --> 01:01:35,280 we're gonna be recording on the yacht from now on yeah yeah uh no but in all seriousness so we 660 01:01:35,280 --> 01:01:40,400 have a you can sign up newsletter dot monkeypatching.io there is no newsletter yet 661 01:01:40,400 --> 01:01:45,600 but we are definitely playing with a bit with how can we make something that is um 662 01:01:45,600 --> 01:01:53,840 quickly digestible trusting something that you can um like do a look a five-minute read or 663 01:01:53,840 --> 01:01:57,920 maybe not even five minutes like once a week where that gives you a bit of an overview what happened 664 01:01:57,920 --> 01:02:02,720 this week a bit like the the the quote-unquote paper version of what we're doing here i think 665 01:02:02,720 --> 01:02:07,600 that might be interesting for some people also use the channel to explain a little bit more on 666 01:02:07,600 --> 01:02:13,600 what we are gonna do with some ideas to repackage we are playing with some ideas around events 667 01:02:13,600 --> 01:02:22,160 we're also playing with some ideas to bring startups young startups in contact with interested investors 668 01:02:22,160 --> 01:02:30,000 like there are a lot of ideas that we're working on but that have not been fully made concrete but if 669 01:02:30,000 --> 01:02:35,680 you want to stay up to date do sign up on a newsletter dot monkey patching dot io also we're gonna have 670 01:02:35,680 --> 01:02:40,640 to think a bit like how to make this valuable right there's a lot of ai generated content and i think 671 01:02:40,640 --> 01:02:45,920 we're both very much like okay if that's how it's gonna be then let's not do it right but i also do 672 01:02:45,920 --> 01:02:50,800 think that having like a different format on this even if it's the same content sometimes like i hear 673 01:02:50,800 --> 01:02:55,360 something but then i'm like okay i need to sit down and read this because just hearing it like it's harder 674 01:02:55,360 --> 01:02:58,960 like sometimes you'd also like to really dive deep or to you have some you need some visuals you need 675 01:02:58,960 --> 01:03:02,800 this so we'll definitely have a thing we don't know if it's going to be all the articles or just 676 01:03:02,800 --> 01:03:07,440 going to be the highlights or just going to be this just going to be that but uh yeah feel free to 677 01:03:07,440 --> 01:03:12,320 to subscribe if you like to stay up to date if you also have any any thoughts suggestions as well 678 01:03:12,320 --> 01:03:19,200 feel free to to reach out there are also contact details of uh our general one mine and merilos on 679 01:03:19,200 --> 01:03:24,880 monkeypatching.io also even for the the regular quote-unquote podcast right like if you have any 680 01:03:24,880 --> 01:03:29,440 article that you've seen throughout the week that you love that you hate it that you would like to to 681 01:03:29,440 --> 01:03:36,320 hear us discussing about it feel free to to let us know and uh we'll be super happy yeah we also did 682 01:03:36,320 --> 01:03:43,920 a short write-up today i published it on my blog bartz.space i also cross post on linkedin on a bit 683 01:03:43,920 --> 01:03:49,920 of uh like we're six months in a bit on the stats that we have so far a bit on these plans that we have 684 01:03:49,920 --> 01:03:56,320 coming up um so if you're interested uh check it out indeed i think this is also a good um 685 01:03:56,320 --> 01:04:02,560 like we're saying this now but i think it's also it's another one of those good like to digest it for 686 01:04:02,560 --> 01:04:07,600 like we sat down you wrote it like you sat down you wrote about it so if you want to to read a bit 687 01:04:07,600 --> 01:04:11,280 more in detail and not just the high level discussion i think it's also something something 688 01:04:11,280 --> 01:04:16,160 something interesting indeed and again have any thoughts feedback let us know we'll be super happy 689 01:04:16,160 --> 01:04:22,560 here and um always very free to leave a five star review just five nothing below five unless you're 690 01:04:22,560 --> 01:04:28,480 unless you unless your scale goes to 10 then no but yeah leave a if you leave a review star we'll be 691 01:04:28,480 --> 01:04:33,760 super super happy as well thanks everybody for listening thanks everyone we'll see you uh next week 692 01:04:33,760 --> 01:04:40,960 yes thank you bart maybe we'll have more decor decoration decor maybe around we can maybe buy uh from all the 693 01:04:40,960 --> 01:04:48,000 the data that we sell like a van gogh maybe issue like maybe be a monkey like yeah you know figure it out 694 01:04:48,000 --> 01:04:51,600 thanks everyone ciao