WEBVTT 00:00:12.280 --> 00:00:16.139 Welcome to the Bikeshed Podcast, the show where we talk all things software engineering, 00:00:16.820 --> 00:00:21.480 agree on a topic ahead of time, and then immediately abandon it, just as the name suggests. 00:00:22.140 --> 00:00:26.980 I'm your co-host, a man who just cooked up his next feature with Claude, ran format on save, 00:00:27.360 --> 00:00:29.640 and confidently called it his own, Scott Kaye. 00:00:30.520 --> 00:00:32.640 And alongside with me are my co-hosts. 00:00:33.340 --> 00:00:39.160 Even during his vacation, his GitHub contribution graph lights up like the brightest house on the block at Christmas. 00:00:40.060 --> 00:00:43.820 He's the reason your mom has 47 items in her online shopping cart. 00:00:44.560 --> 00:00:47.620 He's not responsible for the 45-second My Account load time. 00:00:48.360 --> 00:00:53.100 And if your company just migrated to his open source tools, that load time might actually be reasonable. 00:00:53.880 --> 00:00:54.460 Matt Hamlin. 00:00:55.510 --> 00:00:56.440 And my other co-host. 00:00:57.300 --> 00:01:02.840 A man whose takes are so hot that Cloudflare pays him a retainer just to keep their CDN cash warm. 00:01:03.820 --> 00:01:07.360 You may have heard he recently brought down Verizon by simply asking, 00:01:08.200 --> 00:01:09.460 yeah, but why though? 00:01:10.380 --> 00:01:12.380 Dillon, spicy, take, Curry. 00:01:13.500 --> 00:01:15.920 Matt, Dillon, what's on the agenda today? 00:01:18.020 --> 00:01:21.300 This week we're going to cover some of our 2026 predictions. 00:01:22.060 --> 00:01:26.940 Yeah, the three of us spent maybe the past week outlining some predictions 00:01:26.960 --> 00:01:31.720 that we have for the year or some you know maybe hopeful predictions maybe pessimistic predictions 00:01:32.320 --> 00:01:37.140 that we expect to see pan out over the next year so yeah so without further ado i guess like 00:01:37.660 --> 00:01:41.260 let's dig in Dillon do you want to share your first prediction 00:01:43.240 --> 00:01:49.440 Sure we're probably going to have some like overlap but let's see what happens um all right my first 00:01:49.490 --> 00:01:56.920 one is and i feel like if you work at a certain place this might already exist but autonomous 00:01:56.960 --> 00:02:02.580 coding agents are going to find their way into our workflows single prompt to deploy that's how 00:02:02.580 --> 00:02:09.220 i'm thinking about it so i'll just like not have to write code it anymore and it will just like 00:02:09.220 --> 00:02:09.220 the entire flow will be automated 00:02:09.221 --> 00:02:09.221 Are you imagining this like getting more widely adopted because i 00:02:16.100 --> 00:02:22.400 feel like it's gotten like for smaller sort of side projects things like v0 and bolt have like 00:02:23.000 --> 00:02:29.860 done decently well at some of this but maybe you're thinking of like or you know i guess like 00:02:29.890 --> 00:02:33.160 some of those tools it's like you need to prompt it a decent amount like you have to go through a 00:02:33.160 --> 00:02:33.160 few iterations before you're like like actually have something ready to go 00:02:33.161 --> 00:02:33.161 yeah i think of it as 00:02:39.360 --> 00:02:47.640 more like teams will product teams will start to just write new features using this without an 00:02:47.660 --> 00:02:47.660 engineer and they'll just let it cook 00:02:47.661 --> 00:02:47.661 i hope not actually i hope so because then when it's when 00:02:55.860 --> 00:03:02.780 it's absolute crap and the enshittification of the web gets worse the companies that realize they 00:03:02.960 --> 00:03:07.320 still need their engineers and they still need to think through problems those are the companies 00:03:07.560 --> 00:03:11.320 that we're going to see pull to the forefront this was actually technically one of my predictions 00:03:12.320 --> 00:03:13.840 but here it's now to take. 00:03:14.440 --> 00:03:15.960 Since I came armed with predictions, 00:03:16.240 --> 00:03:18.060 I'm strapped with predictions. 00:03:18.330 --> 00:03:19.380 I don't need this many predictions. 00:03:19.970 --> 00:03:20.700 I'm just going to say, 00:03:22.380 --> 00:03:24.420 it's kind of a continuation of what you're saying, Dillon, 00:03:24.640 --> 00:03:29.160 but I think basically that companies 00:03:29.560 --> 00:03:32.340 that we're going to see in 2026, 00:03:32.580 --> 00:03:36.300 companies that actually care about the end results 00:03:36.580 --> 00:03:38.580 start to pull forward in front of other companies. 00:03:38.900 --> 00:03:41.120 We talk a lot on this podcast about enshittification, 00:03:41.820 --> 00:03:51.120 I think in 2026, this is the year we're going to see which companies actually put more effort in their products than just we use AI. 00:03:53.459 --> 00:03:56.240 In terms of likeliness, maybe we'll do it on a scale of 10. 00:03:56.460 --> 00:03:56.860 I don't know. 00:03:57.260 --> 00:04:00.500 Like 10 being high, like it is absolute truth that's going to happen. 00:04:00.580 --> 00:04:02.640 And one is like probably never going to happen. 00:04:03.520 --> 00:04:05.240 I think we'll definitely see companies doing that. 00:04:06.040 --> 00:04:10.540 And so from that lens, like I think it's like maybe a six or a seven. 00:04:11.560 --> 00:04:15.840 in my mind. Scott, what's your rating on Dillon's prediction? 00:04:17.959 --> 00:04:22.680 I'm thinking like five. I think because I think this is like going to be like a 50-50. It's very 00:04:22.960 --> 00:04:27.480 similar to what you're saying, Matt. I think people will take those risks. But if we're using like 00:04:28.160 --> 00:04:35.439 my take into reacting to Dillon's, I think we're going to see that absolutely backfire for some 00:04:35.460 --> 00:04:40.840 companies and we're going to see other companies who don't just go off like they might go all in 00:04:40.850 --> 00:04:45.380 on tools but i think we're going to start to see like what are what are the things beyond the code 00:04:45.420 --> 00:04:50.160 is now cheap what are the things beyond the cheap code that now make an app great and i think we're 00:04:50.160 --> 00:04:55.020 going to start to see what companies come to the forefront for that so i'd say it's like a five on 00:04:55.050 --> 00:05:01.639 the sense it's it's very realistic but like when it comes to like we didn't go on the next meter 00:05:01.660 --> 00:05:02.840 here, but I think this is 00:05:03.780 --> 00:05:05.600 I think this is like an up 00:05:05.680 --> 00:05:07.460 there take like it. What are the 00:05:07.680 --> 00:05:09.380 options? I think Diablo is the hottest 00:05:09.780 --> 00:05:11.620 take, right? Or do 00:05:11.620 --> 00:05:13.700 we have to hold off on the packets? I think it's a fire 00:05:13.900 --> 00:05:15.300 take. So it would be Diablo 00:05:16.000 --> 00:05:17.740 fire, hot 00:05:18.460 --> 00:05:18.800 mild 00:05:19.919 --> 00:05:21.780 Verde, right? Something like that. 00:05:22.140 --> 00:05:23.700 Anyway, I think we're in a fire 00:05:23.880 --> 00:05:25.040 take zone. Like this is a 00:05:26.320 --> 00:05:27.640 good strong take here. 00:05:28.420 --> 00:05:29.620 And I do think it'll happen. 00:05:29.970 --> 00:05:31.620 I just think there's going to be 00:05:31.640 --> 00:05:33.600 I mean, if someone's super successful with it, that'd be great. 00:05:33.800 --> 00:05:37.820 But I think there might be repercussions for some who use AI in a lazy way. 00:05:39.540 --> 00:05:44.080 Yeah, it's more of what I've seen more recently is there's just more and more MCPs. 00:05:44.190 --> 00:05:48.840 And then there's like more and more of using agents to invoke other agents. 00:05:49.180 --> 00:05:56.920 And it's just becoming this like, I don't know, storm of like agents will be able to control agents. 00:05:57.520 --> 00:06:02.860 And eventually it'll be able to like autonomously drive itself in a sense. 00:06:04.160 --> 00:06:07.240 Well, I think what makes the take fire, I actually think that is the future. 00:06:07.940 --> 00:06:10.140 Now, how long it takes to get there is hard to predict. 00:06:11.900 --> 00:06:17.100 But I think what makes the take to me like fire is you're saying like we don't really might not need engineers this fast. 00:06:17.500 --> 00:06:29.880 I think personally that there's probably always going to be at least one engineer that has to now orchestrate all the agents together and understand it. 00:06:30.000 --> 00:06:32.320 It seems like you think like that might become. 00:06:32.740 --> 00:06:37.240 Yeah, that's why I think your take is fire, because you're saying like, oh, we won't even need that. 00:06:37.380 --> 00:06:42.360 I think that the engineer is going to need to be around for longer than longer than this year. 00:06:42.660 --> 00:06:44.100 But yeah, I don't know. 00:06:44.400 --> 00:06:44.480 Yes. 00:06:44.680 --> 00:06:44.760 Yeah. 00:06:44.980 --> 00:06:50.260 If it happens this year, it'll be surprising, but I think it's possible because I'm starting to go away from it. 00:06:50.780 --> 00:06:50.980 Lean in. 00:06:51.050 --> 00:06:58.940 Yeah, I've seen where they're creating sandboxes where you can just like create 10 different versions of your feature in a sandbox and then you can review it. 00:06:59.310 --> 00:07:02.420 I feel like a product person can just review it and approve it and just move on. 00:07:03.460 --> 00:07:05.480 And like, meet us. 00:07:06.620 --> 00:07:07.940 It kind of reminds me of a cursor. 00:07:08.200 --> 00:07:10.540 Cursor is now like the designer's dream 00:07:10.990 --> 00:07:13.380 because like cursor can just basically do everything 00:07:13.600 --> 00:07:14.940 and the designer can design it. 00:07:15.460 --> 00:07:17.380 And it's like, we don't even need you, engineer. 00:07:18.600 --> 00:07:20.360 I shared this, there was this tweet 00:07:20.820 --> 00:07:21.480 that I saw the other day 00:07:21.480 --> 00:07:23.640 and I shared it with one of our Discord servers. 00:07:23.960 --> 00:07:28.060 But this person's like, so for those who don't know, 00:07:28.120 --> 00:07:29.720 there's this thing called Gastown, 00:07:30.020 --> 00:07:33.460 which is a orchestrator for AI agents. 00:07:34.600 --> 00:07:36.420 You have like multiple tiers of like a, 00:07:36.700 --> 00:07:39.980 You have a mayor that manages pole cats, 00:07:40.900 --> 00:07:42.120 which do the work. 00:07:42.280 --> 00:07:43.300 And then there's deacons that, 00:07:43.660 --> 00:07:44.000 I don't know. 00:07:44.140 --> 00:07:44.680 There's a lot of characters, 00:07:45.260 --> 00:07:46.460 but gas town, 00:07:46.740 --> 00:07:47.040 this thing, 00:07:47.380 --> 00:07:48.620 it's like an orchestrator for multiple agents. 00:07:49.220 --> 00:07:49.740 And this tweet, 00:07:50.000 --> 00:07:50.560 this guy's like, 00:07:50.960 --> 00:07:51.140 well, 00:07:51.140 --> 00:07:53.520 I started sparing up multiple gas towns to work on it. 00:07:53.980 --> 00:07:54.200 It's like, 00:07:54.300 --> 00:07:55.340 there's multiple town instances. 00:07:55.660 --> 00:07:57.880 Now the towns are competing in terms of work. 00:07:57.940 --> 00:07:58.380 And so now, 00:07:59.000 --> 00:08:02.160 now there's soldier agents that go and fight their towns. 00:08:06.819 --> 00:08:14.180 i can just imagine that like scaling out to gas countries and gas worlds uh by the end of 2026 00:08:16.020 --> 00:08:16.020 and there's no role for us anyway um my 00:08:16.021 --> 00:08:16.021 very possible 00:08:16.022 --> 00:08:16.022 my my spice uh 00:08:26.440 --> 00:08:32.919 or my spice rating for Dillon's first prediction uh i think it's in between fire and verde i think 00:08:33.419 --> 00:08:33.419 do we have is there a spice there i'm trying to remember the scale um 00:08:33.420 --> 00:08:33.420 verde is the lowest right 00:08:38.680 --> 00:08:38.680 i think the scale i'm sorry i'm gonna try diablo fire hot mild verde 00:08:38.681 --> 00:08:38.681 okay i think we should also 00:08:50.840 --> 00:08:52.720 throw in a white bread or a mayo 00:08:53.580 --> 00:08:54.080 below Verde. 00:08:55.820 --> 00:08:57.160 Not even on the scale. 00:08:57.300 --> 00:08:57.420 Okay. 00:08:59.120 --> 00:08:59.600 Yeah. 00:08:59.600 --> 00:09:01.100 I guess it's a hot take 00:09:01.140 --> 00:09:02.240 in my opinion. 00:09:05.000 --> 00:09:06.120 I don't think we're going to hit 00:09:06.260 --> 00:09:08.740 in the same way that I think some orgs will do it 00:09:08.740 --> 00:09:09.360 and some won't. 00:09:09.740 --> 00:09:12.360 I think from what I've been seeing 00:09:12.540 --> 00:09:14.420 in terms of AI development 00:09:14.520 --> 00:09:16.720 in the past six months, past year, past two years, 00:09:17.300 --> 00:09:18.900 I could see this 00:09:19.680 --> 00:09:20.060 taking off. 00:09:21.140 --> 00:09:22.940 Dillon's saying like engineers won't be needed. 00:09:23.210 --> 00:09:25.380 I don't know if anyone's going to go that far this year. 00:09:25.570 --> 00:09:28.360 I think that is a really, that's what makes it so far. 00:09:30.180 --> 00:09:30.360 All right. 00:09:30.390 --> 00:09:34.460 So I think we then also need to review, Scott, you shared your first prediction. 00:09:35.500 --> 00:09:37.100 Can you reiterate it for the listener? 00:09:37.490 --> 00:09:38.920 And then we can give a review of it. 00:09:39.060 --> 00:09:39.980 I'm not going to count that. 00:09:40.040 --> 00:09:40.900 I'm just going to use that. 00:09:41.060 --> 00:09:41.960 So I have a few predictions. 00:09:42.270 --> 00:09:44.920 We can, I can just use that as some of my logic. 00:09:45.330 --> 00:09:49.519 I do want to just like maybe cover it one quicker time that, that there's going to be some sort 00:09:49.540 --> 00:09:53.620 of reward or merit for companies that care about avoiding enshittification. 00:09:54.720 --> 00:09:57.820 And we'll see those companies emerge in 2026. 00:09:58.380 --> 00:10:05.520 Maybe not like super hot take, but basically we're going to see leaders in how to use AI 00:10:05.820 --> 00:10:06.000 correctly. 00:10:06.080 --> 00:10:07.520 And that's going to actually emerge. 00:10:07.760 --> 00:10:13.180 Like they're going to become the gold standard for what should an engineering team do with 00:10:13.240 --> 00:10:14.700 AI and how should it be used? 00:10:15.160 --> 00:10:16.640 Similar to, I guess, Gastown. 00:10:17.160 --> 00:10:23.880 But hopefully it's not just like these Ralph Wiggum Gastown solutions for agents. 00:10:24.280 --> 00:10:26.320 Haven't you heard that Ralph Wiggum is out? 00:10:26.720 --> 00:10:28.340 The Cloud Code shipped tasks? 00:10:28.760 --> 00:10:29.580 It's already out. 00:10:30.360 --> 00:10:30.880 And it's out. 00:10:30.980 --> 00:10:32.700 Yeah, it's been around for two weeks and it's gone. 00:10:34.140 --> 00:10:38.680 Boomers are using Ralph Wiggum loops, but all the Gen Zs have moved on to, 00:10:38.960 --> 00:10:41.060 or Gen Alphas have moved on to something better. 00:10:41.600 --> 00:10:44.060 Don't hate me, but I never used a Ralph Wiggum loop. 00:10:44.180 --> 00:10:48.180 I mean, I use many agents, but not a Ralph Wiggum loop. 00:10:48.480 --> 00:10:49.100 I don't know. 00:10:49.940 --> 00:10:51.960 I'm still relevant in 2026. 00:10:52.600 --> 00:10:53.760 That's what I'm trying to do. 00:10:54.740 --> 00:10:56.040 Okay, so my prediction. 00:10:57.100 --> 00:10:58.900 My prediction, first prediction for 2026. 00:11:00.100 --> 00:11:04.400 AI in interviews for engineers will become the norm. 00:11:05.230 --> 00:11:10.619 And that means that engineers will be using AI in their interview process 00:11:11.220 --> 00:11:14.440 in the coding round as the norm. 00:11:16.260 --> 00:11:17.780 I guess I could lean in on this. 00:11:19.080 --> 00:11:22.660 It's pretty wild to me that you ask somebody 00:11:22.970 --> 00:11:24.960 how to use something that makes your job easier, 00:11:25.660 --> 00:11:27.760 but then you would interview them without it. 00:11:28.050 --> 00:11:30.580 But not only you ask them how to use it, 00:11:30.850 --> 00:11:33.040 to use it, you force it down their throats 00:11:33.210 --> 00:11:34.400 in a lot of these companies. 00:11:34.620 --> 00:11:35.580 A lot of these companies are like, 00:11:35.990 --> 00:11:38.040 you know, tokens, we don't care about tokens. 00:11:38.340 --> 00:11:39.380 Burn them up, burn them up. 00:11:39.520 --> 00:11:41.840 You don't want to do any coding anymore. 00:11:42.020 --> 00:11:43.240 Just burn them tokens up. 00:11:43.580 --> 00:11:48.760 So you would rather spend more money using AI to solve problems, 00:11:49.480 --> 00:11:51.460 but hiring people who are smart enough to do something. 00:11:51.580 --> 00:11:53.120 Well, if it's actually part of the workflow, 00:11:53.160 --> 00:11:56.060 we're going to see reasonable companies start to understand this 00:11:56.520 --> 00:11:59.280 and try to find and learn from how people use AI 00:11:59.880 --> 00:12:02.560 to try to come up with some of these emerging behaviors 00:12:02.920 --> 00:12:06.300 that companies might start to see and trends we might start to see. 00:12:07.080 --> 00:12:08.900 Like Ralph Wiggum and Gastown. 00:12:09.780 --> 00:12:14.700 any immediate reactions Dillon or do you want to jump into your uh you know your rating of how 00:12:14.820 --> 00:12:14.820 likely 00:12:14.821 --> 00:12:14.821 the place i work at is already telling people to use ai during interviews like that's 00:12:19.860 --> 00:12:27.480 a part of our prompt which is weird but the reason i say it's weird is i don't think our questions 00:12:28.000 --> 00:12:33.839 are written in a way that makes them work well with ai and i think that's going to be the biggest 00:12:33.860 --> 00:12:42.340 problem with asking people to use ai if we don't rewrite our problems to to work well with ai then 00:12:42.860 --> 00:12:48.200 we're just going to get really really bad candidates but yeah i think this is very likely to be true 00:12:48.250 --> 00:12:48.250 because i'm already seeing it and then the spice level is what's above mild hot 00:12:48.251 --> 00:12:48.251 it was uh diablo 00:12:57.740 --> 00:12:59.920 low fire, hot, mild, verde mayo. 00:13:00.460 --> 00:13:02.040 It's in between fire and hot for me. 00:13:03.220 --> 00:13:03.880 I think that's fair. 00:13:04.560 --> 00:13:05.420 It's not a hot take. 00:13:05.630 --> 00:13:07.040 I mean, I want to just call out too. 00:13:07.220 --> 00:13:08.460 Yeah, we did this a total bit as well. 00:13:08.850 --> 00:13:12.140 And I think startups usually catch on to some of these trends faster 00:13:12.370 --> 00:13:15.280 because, you know, they just need people who can move quickly. 00:13:15.430 --> 00:13:17.000 So this is a great way to move faster. 00:13:17.820 --> 00:13:21.480 So smaller companies are probably quicker to take on this. 00:13:22.380 --> 00:13:23.460 I guess I was going to call it a risk. 00:13:23.530 --> 00:13:26.540 I wouldn't say necessarily it is a risk, 00:13:26.780 --> 00:13:31.160 But I meant to also just elevate that like bigger corporate companies, right? 00:13:31.280 --> 00:13:34.800 Like we're going to see Meta, we're going to see Google, we're going to see Amazon starting 00:13:34.920 --> 00:13:35.460 to allow it. 00:13:35.880 --> 00:13:37.360 So view it through that lens. 00:13:38.160 --> 00:13:41.800 Yeah, my prediction in terms of how likely this is true. 00:13:41.990 --> 00:13:43.160 Yeah, I think nine out of 10. 00:13:43.230 --> 00:13:47.840 I think it's, you know, I'm sure there's already companies doing this, but I think it's, I 00:13:47.850 --> 00:13:51.640 think we're going to see a lot of companies that maybe took the hard stance a year ago, 00:13:51.880 --> 00:13:58.820 six months ago two years ago to say like no ai use at all during the interview to lean into it 00:13:59.140 --> 00:14:03.960 and instead like sort of flip that decision a little bit um and then in terms of like the 00:14:04.080 --> 00:14:08.160 spiciness yeah i think it's i think it's actually pretty lukewarm i think i think probably a fair 00:14:08.160 --> 00:14:13.180 verde for me is like i think it's there's going to be people that are going to react spicy like 00:14:13.480 --> 00:14:18.959 obviously are going to have hot takes on it but i think just like i think it's bound to happen 00:14:19.540 --> 00:14:23.940 it's like it's like telling someone it's like no you can't write your uh you know you can't use 00:14:24.100 --> 00:14:29.320 javascript in your interview you have to use c or c plus plus it's like no that that's silly 00:14:30.220 --> 00:14:35.320 cool i totally think that's fair matt why don't you give us your first prediction of 2026 00:14:36.600 --> 00:14:43.220 all right let's see um i guess i'll start with this one i think this is maybe actually kind of 00:14:43.420 --> 00:14:48.940 counter to what our the year is first predictions were but i think companies will actually start to 00:14:48.960 --> 00:14:54.460 back on their AI budgets. I think they've been throwing a lot of money at it over the past year 00:14:54.560 --> 00:14:59.980 and a half year, two years, whatever it is. Um, and I think we're going to start to see companies 00:15:00.120 --> 00:15:04.900 realize that actually, no, we don't need to pay for, you know, both a Claude subscription for every 00:15:05.100 --> 00:15:09.680 seat and a cursor subscription for every seat and a GitHub copilot subscription for every seat 00:15:10.360 --> 00:15:15.340 and chat, like all of these different services that overlap quite a bit. I think we're going to 00:15:15.240 --> 00:15:21.340 start to see these companies start to say, actually, we don't want to spend that much 00:15:21.440 --> 00:15:21.920 money on AI. 00:15:23.120 --> 00:15:28.860 Let's try to centralize our expenses a little bit and try not to just, I don't know, we're 00:15:28.920 --> 00:15:34.000 not in a zero interest rate sort of economy anymore. 00:15:34.200 --> 00:15:36.300 So I think that's definitely going to put additional pressure there. 00:15:38.259 --> 00:15:43.420 Do you think that will reduce how much AI we're using or we're just going to pick a winner 00:15:43.860 --> 00:15:46.140 for a company that we're using AI with? 00:15:47.220 --> 00:15:49.320 I think it could be a little bit of a mix of both. 00:15:49.390 --> 00:15:51.880 I think you could see some companies saying, 00:15:52.210 --> 00:15:55.520 okay, we've been paying for maybe in the coding agent space, 00:15:55.680 --> 00:15:58.120 we've been paying for Claude and for ChatGPT, for example, 00:15:58.300 --> 00:16:01.000 or OpenAI's offerings. 00:16:01.720 --> 00:16:03.040 I think we're going to see some companies say, 00:16:03.120 --> 00:16:04.500 actually, we don't need to be paying for both, 00:16:04.530 --> 00:16:06.300 and we're just going to go in on one. 00:16:07.440 --> 00:16:10.360 But I think also you could see a little bit of, 00:16:11.860 --> 00:16:15.840 Yeah, so that's sort of condensing down into what provider you use. 00:16:16.110 --> 00:16:17.360 I think we'll also just generally see, 00:16:18.150 --> 00:16:20.600 actually we don't need to be paying for, I don't know, 00:16:21.250 --> 00:16:24.180 10,000 seats for this application 00:16:24.410 --> 00:16:26.440 when only 5,000 people are actively using it. 00:16:28.040 --> 00:16:29.440 Again, random numbers pulled out of a hat. 00:16:29.630 --> 00:16:30.960 But I think we'll start to say, 00:16:31.540 --> 00:16:35.080 be a little bit more careful about just throwing money at AI. 00:16:36.080 --> 00:16:36.680 Wow, okay. 00:16:37.420 --> 00:16:39.600 Dillon, do you have any more reaction before I go ahead? 00:16:40.000 --> 00:16:41.420 You also need to rank it. 00:16:42.040 --> 00:16:42.540 you go first 00:16:44.160 --> 00:16:50.740 okay so this one is very similar to another one of mine um it's basically the same thing um 00:16:51.500 --> 00:16:55.120 and the reason why i picked it is i feel like it's slightly a double-edged sword 00:16:55.600 --> 00:16:59.000 i think you're right in the sense that like some companies will cut back we know 00:16:59.700 --> 00:17:05.780 at a former company we worked at they don't have a lot of tokens which is not fun uh but anyway 00:17:06.449 --> 00:17:11.400 i think like the magnificent seven companies since they're like really been propping up the 00:17:11.420 --> 00:17:16.760 industry they'll continue really big companies will continue to burn tons and tons of money on 00:17:16.850 --> 00:17:21.819 ai because they see it as a money maker however i do agree with you in the sense that maybe some 00:17:22.420 --> 00:17:28.760 other companies would cut back i guess my take was more of like people will start to think ai 00:17:29.140 --> 00:17:35.800 is super helpful so ai prices will rise and will care more about what you burn tokens on 00:17:36.320 --> 00:17:38.500 I don't want to use it anymore because it's very similar. 00:17:40.620 --> 00:17:42.960 But again, I think it's like a double-edged sword. 00:17:43.090 --> 00:17:55.380 I think that, yes, some companies might start to scale back to see where they can make savings because it is crazy how we're throwing money at the AI bubble, I guess, right? 00:17:56.980 --> 00:18:01.260 So thoughts are mainly, I think this is on a scale of 1 to 10. 00:18:02.040 --> 00:18:05.780 I think this is very likely a five or a six again. 00:18:06.220 --> 00:18:07.120 Maybe we'll just say a six. 00:18:08.520 --> 00:18:16.100 But it's a double-edged sword in the sense that basically companies that have the money are going to continue to just throw money at the wall. 00:18:17.720 --> 00:18:24.060 And companies that are more budgeted that can't really afford a black hole won't be able to handle that. 00:18:24.580 --> 00:18:29.600 So on the take scale, I'm going to give this a mild. 00:18:30.080 --> 00:18:31.620 I'm going to say that's better than it sounds. 00:18:32.300 --> 00:18:38.440 it's not a white bread or a or a verde take it's mild but i i think it's reasonable and logical 00:18:38.920 --> 00:18:38.920 for these companies 00:18:38.921 --> 00:18:38.921 um i think this is tomato sauce level of spice 00:18:46.320 --> 00:18:48.960 in that in that you think it is pretty likely to happen or 00:18:49.720 --> 00:18:49.720 just not likely to happen and it's not very spicy 00:18:49.721 --> 00:18:49.721 oh that's surprising to me 00:18:56.600 --> 00:19:01.800 yeah the read read like re give me your take and let me give you like a good assessment of it 00:19:01.820 --> 00:19:01.820 quick 00:19:01.821 --> 00:19:01.821 uh my take is just that companies will start to scale back on their expenses for ai i'm not 00:19:09.940 --> 00:19:15.300 necessarily saying that they're going to flip in the sense that they start to start to like not 00:19:16.000 --> 00:19:16.000 spend money on it but just like spend less 00:19:16.001 --> 00:19:16.001 i still think this is the year i i feel like the ai train 00:19:22.850 --> 00:19:28.840 is going to continue to move at full speed ahead for the entire year i feel like there's just so 00:19:28.860 --> 00:19:36.360 many new things emerging out of it right now um that it's just not going to slow down and maybe 00:19:36.460 --> 00:19:41.340 that's just where i am but place i'm at the investment is huge and i don't think it's going 00:19:41.520 --> 00:19:41.520 anywhere anytime soon 00:19:41.521 --> 00:19:41.521 so you're also you're also kind of just like maybe wishful thinking or hopeful 00:19:46.180 --> 00:19:50.980 thinking that uh that it does that my prediction is false and so your company keeps spending more 00:19:51.000 --> 00:19:51.000 and you get to keep using those keep consuming those tokens 00:19:51.001 --> 00:19:51.001 i just don't think it's going anywhere 00:19:57.860 --> 00:20:01.040 Our company just started the AI coalition of Massachusetts. 00:20:03.400 --> 00:20:06.120 They're trying to tie it into politics or something at this point. 00:20:06.899 --> 00:20:08.080 They're not slowing it down. 00:20:10.520 --> 00:20:12.040 I'm going to have to put that in the show notes. 00:20:13.700 --> 00:20:16.900 Who's up next? I think we make Dillon do it next again. 00:20:17.140 --> 00:20:18.620 Dillon, what's your second prediction? 00:20:19.600 --> 00:20:24.800 I'm going to try to steer it away from AI since we all gave an AI take in 2026. 00:20:26.340 --> 00:20:27.860 and maybe give you 00:20:27.980 --> 00:20:29.740 a sillier take 00:20:30.140 --> 00:20:30.580 this time around. 00:20:32.920 --> 00:20:34.120 One of us is going to get 00:20:34.120 --> 00:20:34.620 a new job 00:20:36.380 --> 00:20:38.020 and one of us is going to get laid off. 00:20:38.340 --> 00:20:39.820 Dude, that's literally one of my takes. 00:20:40.560 --> 00:20:41.240 Are we talking about 00:20:42.090 --> 00:20:44.060 the listeners included or just the three 00:20:44.070 --> 00:20:44.070 of us? 00:20:44.071 --> 00:20:44.071 Just us in 2026. 00:20:47.060 --> 00:20:48.080 And it could be the same person. 00:20:48.460 --> 00:20:48.460 Just us. 00:20:48.461 --> 00:20:48.461 It could be the same person. 00:20:50.080 --> 00:20:52.460 It could be the same person. 00:20:56.980 --> 00:21:01.700 i think i'm out of takes i thought i came with so many takes every take you guys have had i've 00:21:01.700 --> 00:21:01.700 had the same take 00:21:01.701 --> 00:21:01.701 so Scott and it's it sounds like you're it sounds like you're 15 out of 10 00:21:06.260 --> 00:21:06.260 on how likely this prediction is going to be true then 00:21:06.261 --> 00:21:06.261 one of my takes was one of us will get laid 00:21:12.420 --> 00:21:18.180 off i didn't even predict we'd find a job i didn't even get that far so Dillon's done a lot more work 00:21:18.200 --> 00:21:18.200 there now 00:21:18.201 --> 00:21:18.201 that's okay 00:21:18.202 --> 00:21:18.202 damn i i should we go back to the ai takes or is this is this what we should 00:21:26.720 --> 00:21:34.360 all right so you're 10 out of 10 on how likely and your spice meter sounds like maybe mayo or like 00:21:34.720 --> 00:21:34.720 like in the sense that you're completely aligned with Dillon's take 00:21:34.721 --> 00:21:34.721 wait which part someone will 00:21:40.180 --> 00:21:40.180 get laid off 00:21:40.181 --> 00:21:40.181 yeah i actually i'm gonna be honest i i was trying to be a little more spicy with it 00:21:47.160 --> 00:21:53.840 I don't actually think at three companies we work at, one of us is really going to get laid off. 00:21:54.060 --> 00:21:56.200 So I think it is a little spicy if we do. 00:21:56.560 --> 00:22:03.220 And that would be more along the line of a prediction of, oh, AI is so great and it's so awesome and we don't care about the budget. 00:22:03.360 --> 00:22:06.420 So let's just take less engineers is actually like what happens. 00:22:07.480 --> 00:22:10.100 I think some companies will do that. 00:22:10.560 --> 00:22:13.440 And I think it's not going to go as well as they think. 00:22:13.500 --> 00:22:20.040 maybe maybe i'm pauliannish one would say about that um i know like twitter kind of did this right 00:22:20.040 --> 00:22:25.520 in what 2021 with when elon musk came in and the app like it is definitely worse but it's still 00:22:25.700 --> 00:22:31.540 serviceable and it's kind of ushered in the whole enshittification of apps i have a feeling that 00:22:31.540 --> 00:22:37.000 we're going to get closer to a place where like some of these apps can't get that shitty i mean 00:22:37.420 --> 00:22:43.500 maybe not amazon and these large conglomerates but if smaller businesses take bigger risks 00:22:44.120 --> 00:22:48.060 on making their app allowing their apps to be worse just to save a little bit of money 00:22:48.580 --> 00:22:50.480 i do think we'll start to see more competition 00:22:52.140 --> 00:22:58.040 gotcha what's your spice spice rating your spice-o-meter rating trademark 00:23:00.440 --> 00:23:06.020 um i'm gonna say it's hot uh what's no i'm gonna give it i'm gonna give them another fire because 00:23:06.800 --> 00:23:08.060 if one of us gets laid off 00:23:08.980 --> 00:23:09.780 I'll just be 00:23:10.240 --> 00:23:10.960 I'm going to be surprised 00:23:11.270 --> 00:23:12.700 the first person I'm going to blame is Dillon 00:23:13.350 --> 00:23:14.000 he gets a fire 00:23:15.760 --> 00:23:16.800 I'm the most likely 00:23:17.520 --> 00:23:18.240 I'm getting a rough 00:23:18.520 --> 00:23:19.680 because I came up with a take 00:23:21.340 --> 00:23:23.280 I feel like it's going to be something 00:23:23.350 --> 00:23:25.320 where I used AI too much at work 00:23:25.540 --> 00:23:27.600 and the quality of my deliverables just goes to shit 00:23:29.640 --> 00:23:31.600 I mean I might be in that bucket 00:23:33.260 --> 00:23:33.960 just say 00:23:34.470 --> 00:23:36.000 I thought you said that's what you wanted 00:23:36.020 --> 00:23:36.020 isn't that what they said they wanted 00:23:36.021 --> 00:23:36.021 and then they'll get the cue clear like did you say fired 00:23:43.220 --> 00:23:49.980 or laid off like just lose our job it doesn't matter if it's part of like a mess or just like 00:23:50.570 --> 00:23:50.570 matters it doesn't matter 00:23:50.571 --> 00:23:50.571 uh no he said laid off it's got to be laid off you can't act actively 00:23:57.860 --> 00:23:57.860 act a fool and get fired that won't come 00:23:57.861 --> 00:23:57.861 well i'm not trying to i'm not trying to make his prediction 00:24:01.980 --> 00:24:01.980 true or at least not half of it 00:24:01.981 --> 00:24:01.981 who knows maybe you are 00:24:01.982 --> 00:24:01.982 uh i think i think it's an 8 out of 10 00:24:09.000 --> 00:24:09.000 for me i think i can see this being true sadly well 00:24:09.001 --> 00:24:09.001 i didn't you didn't let me pick a number 00:24:15.880 --> 00:24:15.880