We Benchmarked Jev, Nimble, & a Simulated Fly Brain Playing DOOM on Matt and Ryan have a chat

Author: Matt Williams · Watch on YouTube

Video authored by the channel shown, not by IANIA; all rights belong to the creator. Transcript and summary generated automatically; the original-language content is preserved.

Original language: English

Original video: We Benchmarked Jev, Nimble, & a Simulated Fly Brain Playing DOOM on Matt and Ryan have a chat · Channel: Matt Williams · 1:07:33 · Watch on YouTube

Topics: AI agents, AI evaluation and benchmarks, Markets and economy, AI models, AI safety

Summary

Matt and Ryan discuss recent AI models and tools, including Jev, Nimble, Muse Spark, and OpenAI’s new Decisions API, focusing on speed, usefulness, privacy, and prompt-injection risks. They compare these decision models with a simulated fruit-fly brain that can play DOOM, and consider the practical and ethical implications of brain simulations. The conversation ranges across tech news, including OpenAI agents, Nvidia OpenShell, patents, and web-scraping services, before moving into personal topics such as gadgets, clothing, music videos, and books. They finish by sharing reading recommendations and debating several books and adaptations.

Key points

The speakers compare AI models they have tested, describing Opus 5.5 as a strong option for their agent-sandboxing work and Muse Spark as fast and useful for broad checklist tasks.
They note that free AI services may retain or use user data, and mention that test runs have exposed credentials, making privacy and access controls important considerations.
Jev and Nimble are discussed as fast decision-making models; the speakers describe possible uses such as filtering prompt-injection attempts or routing requests to larger LLMs.
They caution that small decision models are not automatically immune to prompt injection, and that sub-agents may ignore a separate model’s recommendations.
A simulated male fruit-fly brain that plays DOOM prompts discussion of reward functions, hardware needs, speed, and the ethical questions raised by brain simulations.
The speakers discuss DHH’s remarks about the future of coding, his operating system project, OpenAI’s agent avatars, and OpenAI’s Decisions API.
They critique Nvidia OpenShell as a potentially unnecessary agent-runtime product and recommend looking at native isolation approaches instead.
The conversation touches on a patent dispute involving Apple’s Taptic technology, as well as web-scraping and search tools such as Nimble Way and Super Data.
They share personal interests, including the Commodore 64 Ultimate, an FPGA-based Superstation One, comfortable clothing, and a music-video performance.
The final portion focuses on books, with recommendations including All the Water in the World, The One and Only Ivan, and Alistair Reynolds’s Poseidon’s Children series.

Chapters

Start	Chapter	Summary
00:01	Opening chat and dental updates	Matt and Ryan open with casual banter, then talk about completed root canals and dental problems.
01:35	Trying recent AI models	They discuss OpenAI Dev Day, model versions, and their experiences with Astra, Soul, Fable, and Opus.
04:13	Muse Spark, Jev, and agent workflows	Matt describes using Muse Spark and Jev with OpenCode, including speed, broad repo-maintenance tasks, and data-retention concerns.
07:22	Privacy concerns and alternatives to Jev	They compare Jev’s data-retention policy with Nimble and other options, and discuss how fast these models can respond.
10:01	Decision models and prompt injection	The speakers consider OpenAI’s Decisions API and how small models might screen documents or help route requests while still facing injection risks.
13:09	A simulated fly brain plays DOOM	They explain the virtual fruit-fly brain project and discuss its training, reward function, speed, hardware, and ethical implications.
19:22	DHH, coding, and operating systems	The conversation turns to DHH’s comments about coding’s future, Rails, and his new operating system project.
24:32	OpenAI agents and Space Bunny Alpha	Matt describes OpenAI’s agent avatars and a promising free model demo called Space Bunny Alpha.
27:05	Florida, AI, and a patent dispute	They briefly discuss Florida’s effort to block OpenAI development, then talk about a company’s legal win involving Apple and haptic technology.
34:46	Apple Music Hall and Nvidia OpenShell	The speakers discuss Apple’s music venue and Matt’s criticism of Nvidia OpenShell as an agent runtime.
38:24	Nimble Way and AI search	They distinguish Nimble Way from other companies with the same name and discuss web scraping and search services.
43:02	Canada, Eurovision, and European geography	A tangent about Canada’s Eurovision participation leads into a discussion of EU membership and geography.
47:16	Gadgets, keyboards, and comfortable clothes	Matt talks about wanting a Commodore 64 Ultimate and describes an FPGA project and comfortable joggers.
54:02	Music videos and reading habits	They discuss a memorable awards-show performance and move on to books, including Dungeon Crawler Carl and science fiction.
1:02:51	Book recommendations and wrap-up	Ryan shares recent reading and recommendations, they discuss The One and Only Ivan and other books, then end the conversation.

Quotes

when it’s free are you the product? — 04:44

50 milliseconds. — 09:29

that’s enough to self-drive a car, build a robot — 16:47

The organoid acts like a real human, which is like actually scary. — 19:22

Totally must. must read. — 1:04:57

Concepts and entities

Jev, Nimble, OpenAI, OpenCode, Muse Spark, OpenAI Decisions API, Male CNS, DOOM, NVIDIA OpenShell, Matt Williams

Full transcript (original language)

00:01 Yo. >> Uh, what do I say? >> I figured since I’m in this office, I should like bring a real microphone. >> Okay. When is that happening? [laughter] >> Yeah, I know. It’s only a shirt. I mean, it’s it was an extra one I had lying around. Um, it’s one that N8N sent me. Uh, which was nice of them. >> I I can’t tell. [laughter] And uh yeah. >> Yeah.

00:31 >> So, what’s up with you? >> Well, uh I I thought you’d be happy to know all of my root canals are done and now I get to work on crown. So, uh the pain is gone. >> So, you won’t be sending me photos of teeth anymore for a while. >> No. Or or like some people asked me, no, I also won’t be sending you photos of a new office or like startup that I’ve joined. Like, no. I actually had tooth problems. [laughter] I did have a couple people actually mention to me that the most common >> I think uh is it alibi or excuse to to

01:05 leave work to have interviews or whatever is about this. Uh full disclosure, apparently I’ve got terrible teeth, but yeah, this is not my first nor my second root canal. >> Uh so yeah, I’m just happy they’re over. Yeah, when I got my last uh cavity that was, you know, the the filling was um not it was falling apart because it’ been there for I don’t know 30 years. Um they took it out and they said, “Well, it if if there’s any problems, they’re

01:35 going to have to do a root canal instead.” It’s like, “Oh, there’s no problems.” I mean, and there haven’t been, luckily. >> That is awesome. >> Yeah. So, I’ve got a bunch of news for you, but not too much in terms of like me doing anything with the news. I am so sorry. Uh, today was Open AI Dev Day and and listen, I they said it was available. I I can’t seem to get 6.1 Soul to come up on on my end, so I haven’t tested like the new model on their end. I did spend last week using a lot of Astra until I get kicked off and

02:06 Soul uh after that. I think both of those were like version six. I’m I’m losing track of the version numbers, I gotta be honest, or the model names, especially with OpenAI. I have a hard time keeping up on like Terra, Luna, Soul. I don’t know which is I’m just going by pricing right now and assuming middle of the road is good enough for me because Astro was good, but also um would work for a really long time and then sometimes be very dumb, I guess. Like, and I had it on high as opposed to extra high or max. And apparently that’s a big deal, too.

02:38 >> Regardless, it just wasn’t. And I I’m feeling the same way about Fable 5.1’s all right, but it’s it’s just not good. And and I’ve not had any at level five. Again, a lot of what I’m doing is um agent sandboxing, working on things like lineage of prompts, that kind of stuff. So So things that are are very much similar to some of the work that anthropic may be doing. And it could be that I’m hitting some of those like safeguards or what have you. >> So was 6.0 Astra or 6.1 Astra? 6.1 Soul and I think Astra’s still on

03:11 6.0. >> Okay, >> is my understanding. So >> 6.1 Astra is the one that they said, “Okay, we’re not going to release this because it’s too scary.” >> Well, they said that about 6.0 Astra as well. And then it >> I think two weeks later it came out. >> Wow. We need some money. >> Yeah. I mean, the the real news here is that like honestly I was I was reaching for Opus 5.5 quite a bit. It’s >> I’m not going to say it’s better than 4.6, six, but it seems to be equivalent and cheaper for me, which is what I

03:41 want. >> Uh, and and when I say better, uh, that my test eval or test bed, it’s mostly like BSD stuff, which generally isn’t in the corpus, >> and the agents really love to use bash, and there’s like by default, there’s no bash on BS BSD, and the agents are running on BSD as well, which really confuses them. So, you can put like ash or whatever and and same like bash to it to allow it to work. There’s like a lot of weirdness going on in in my setup, so it’s kind of not fair. But Obus 55 trounces the competition.

04:13 >> Yeah. >> Other than that, I’ve been using something called Muse Spark Free, which I wasn’t really clear on what it was or what it was doing. Um, again, luckily, these are test repos. Nothing too big, >> but some credentials definitely leaked a couple times. Generally data dog credentials. not too scared about it but still these days with app keys like I could make other keys >> it’s a little bit more frightening um I guess all this to say I was using it in open codes then and my understanding of the free models is that they can in fact use those for training so the muse spark

04:44 my understanding is like Facebook just wants all all your data >> and uh if if if you check that check box they’re like well we then we will take it and we will train on this data and then sell it back to you which listen I I totally get it it’s not like other frontier labs aren’t doing the same It is however pretty fast and relatively accurate. >> It’s difficult to complain about free with the speed that you’re getting, especially when I’m launching like 30 agents at a time. >> So, I’m just basically going through all of my repos and saying, “Hey, keep these

05:16 up to date.” That that is one of the other tests that I’ve got. Yeah. >> My repos are generally like read me, markdown, bash, a little bit Python here and there. like nothing too there might be some Terraform you know and it’s it’s got to like update the uh is it providers and then change the syntax do that but it’s not a lot of work it’s just that it’s it’s wider work it’s not very deep but it’s very wide right >> mspark seems to do very well at like that kind of checklist thing in open code at least but I’ve been using something called jev as well to decide

05:49 the not the fate but when the sub agents come back to you’re like I tried to log into that host but it said the known host has changed like I’m I’m stopping here I can’t do anything right it’s like oh okay just just ask Jev what what happens in these cases and open code through Zen provides Jev for free as well and I I would assume the same things apply here what’s interesting though is that the sub aents in all cases I’ve tried so openai claude and open code these sub agents won’t respect a dev decision which is kind of interesting like on the one hand I I get

06:21 it there’s not a lot of logic on the dev side So, it’s really just looking at the context it’s got and then making a decision, right? If it doesn’t know the background context, it could make the wrong decision. Like, it doesn’t know your system prompt or whatever because it’s not it’s not exposed to those when the calls go out. It’ll just make an answer. >> And the way I did it was set up an MCP server that actually calls out to Jez free from the their side because it’s really not the same. You don’t really prompt Jev. You just give it >> one of three types of questions, right? Right. Um, and it’s really fast and really good at the tiebreaker kind of

06:51 stuff that I was using like Tris skills for for earlier just to essentially argue and weigh the benefits. But the sub agents will not respect any of the decision. So if it gets to a host and the host keys changed because I’ve missed like I’ve changed it or changed the OS or what have you, it it will refuse to proceed unless a human intervenes, which isn’t the end of the world. You can tell like your your quote unquote orchestrator to do the things that Jeb’s doing. Anyway, all this to say, open code’s a very good place to test both Jev as well as Muspark if you want to. Caveat mtor I I think is that

07:22 uh yeah, you’re probably giving them training data which >> you know when it’s free are you the product? I don’t know but I suppose that that need to be said you’ve been playing with open >> I’ve been loving Jev recently. >> Okay, nice. Um, and uh, what I don’t love though is that they’re so they’ve they say on their like privacy statement that they will not use your data to train the model, but they will keep your

07:54 data >> perhaps >> until they change the user agreement later [laughter] >> until they decide what to do with it. And they don’t say what they’re going to do with it. >> They say they’re going to keep your data. Um, unless you’re an enterprise customer, in which case you can sign up for not retaining data. And that’s that’s it. I mean, I was like, >> we’re gonna we’re gonna keep it, and when we find something good and profitable to use with your data, great, we’ll use it. Um, so that’s kind of iffy, you know, especially if

08:27 you’re using this as a kind of a front end to process all your or your, you know, other documents. >> Well, this seems like a terrible idea and it would be it would be terrible if there weren’t alternatives, >> right? >> Like Nimble. >> I know of Kev. Oh, Nimble. Okay. I only know of Kev. Um uh what’s another one? Tev one te um and with 0.35

08:59 of lama now supports nimble and te one and maybe others. >> No way. >> Yeah. And the speed is exactly the same as Jev. In fact, with TE one, it’s a smaller model. It’s faster, like 30% faster than Jev. Um, and and it seems to spit out the right answer just as often as Jev does. >> So, >> yeah, I mean, just pretty fast, I think,

09:29 is the main thing. Fast. >> I mean, it’s uh like uh 70 60 70 milliseconds and TEV is often 50 50 milliseconds. >> Um, from >> for those watching, that’s Yeah. Yeah. For those watching and spitting it out. 50 milliseconds. >> Exactly. Um it I I want to make sure we we make this clear because like for most people if I like show something on the screen and you have to move your mouse and click on it, >> it’ll take you 60 milliseconds. So it’s as fast as you can think and drag your mouse basically which is it’s non-trivial. Uh now I did bury the lead

10:01 a little bit because I I I kind of start it wrong but OpenAI as a result launched the decisions API today which I have not also playbook >> but it does look like their answer quote unquote to >> Jev. >> What I find interesting is that >> and and this is kind of like I don’t think it’s it’s out there yet. The creator of Jev said they were going to release a little bit more about the training data, how they built their base. That that has yet to happen. Uh it seems like it might have been Deepseek because it it would answer like who are

10:32 you on Deep Seek R1 or something like that at one point in time. >> Um so it might be that they’ve like forked or basically done something to do one of the deep sea piles to do it >> which I find interesting because I think that’s what Kev did. They’re like, “Oh, I mean I’ll steal your homework and see if I can go down the same path.” And apparently Kev is pretty good. but it’s not as fast and and maybe not as accurate. But I would like to try Tev as well as this other one here. >> Tev and Nimble. Um and uh yeah what the

11:04 things I find interesting about um about about these models is >> um you know with one of the benefits of a you know if you’re processing email or chat messages or whatever it is um and you’re using an open AI or clawed versus a local model the without any proof proof. People are just assuming that the open AI and anthropic models are going

11:36 to be better at rejecting prompt injection attacks >> without true. [laughter] >> Yeah. And even if it were true, um, Jev and, uh, and OpenJV and Nimble and all these others, um, they’re not susceptible to that same kind of they might be as susceptible to injection attacks if the user knows if the attacker knows exactly what the criteria is and all those types of things. And maybe right >> they can and but

12:08 >> um you know now you’ve got this this model that says are there any prompt injection attacks in this email? No. Okay, great. Hand it off to the LLM to process further. um or it’s one other use case I saw was um for understanding you know that that pre that gate that that uh gate for accessing your LLM you know it can read massive documents and and then provide some you know 10

12:38 answers on this document yes or no uh answers um [clears throat] and it’s fast and cheap uh even with Jev and now with Nimble, it’s even cheaper if you’ve got the hardware. Uh then uh you know that that can be that that gateway to your LLMs or decide which LLM to use. Another example I’ve seen others use uh super fast for that kind of that kind of thing. I think it’s super cool. >> Yeah.

13:09 >> Have you seen the Doom the Jev playing Doom? >> Yes. Yes. >> Yes. which I I think all of us could rationalize about but like the programming is not difficult but you need something that can think this way. Yeah. >> Um now have you seen the fly brain that plays Doom as well? >> No. >> Okay. So we we’ve been talking about organoids a bit. I’m not sure if you remember what organoids are. So they’re vat grown brains that you can use to to do things. >> Uh we we have not done this for flies.

13:41 What we’ve done, my understanding is we’ve taken nano slices of the brain and fully modeled it in a virtual environment and shipped that [laughter] model. Um, so I think it’s called uh >> mail CNS is it? >> That is incorrect. That is correctly what it is. Yeah. So So male it’s it’s a male fly. >> [laughter] >> I forget what CNS is, but I think the N is for neural and S is maybe simulation.

14:14 >> Um, yeah, it’s pretty interesting. >> And then you tap the reward function. So, the the dopamine basically based on like I think it’s completing the level in Doom and it might have taken a while to actually quote unquote learn it, but yeah, this is yet another >> another way to uh do it. So, I think Let’s see here. So, it’s mapping a fruit brain. I’m trying to look through. I mean, it’s

14:44 like GitHub. Sorry. Um Minecraft Beat Saber. Oh, parallel park to car. Uh traded Bitcoin. He’s cutting donor kebabs. Um, and all this sounds really funny until you start thinking about the moral implications and it’s like, “Oh, wait. But actually, this is a simulation of a brain.” So, I don’t I don’t even know what to do with this. Like, like my Yeah. jury’s out on like how I feel like this. Like, if Matt, if I let let’s I’m

15:14 so sorry. Let’s assume you pass away tomorrow, but we caught it in time and we slice your brain. Yeah. >> In I don’t know, a million slices. >> Would you be angry? Or would like the million copies of you that result I wouldn’t care, >> right? Like [laughter] should be okay. >> Yeah. >> I mean, I’ve got that thing on my driver’s license says donate my organs to whoever >> including brains. Yeah. Yeah. Yeah. Yeah. And and I I Okay, so this is me thinking about my body and and sorry if

15:46 this becomes part of the public record, but I think I would be okay living on somebody else’s phone and and like being me. Yeah, >> like as a simulation of my brain. >> I think you got the video kind of weird there. You’re you’re smaller than me. Sorry. You’re looking more fly brain than I am. >> That’s weird. I must have I must have dragged something. Let’s see if I can >> nudge something. There we go. Yeah, make me smaller. >> Move me over there. >> Yeah, just just overlay. Just put you over my face would be perfect. >> Let’s see. Make me myself a little bit

16:16 bigger. I wonder what happened there. And then we’ll line you up there. No idea. Look at this guy go. >> If only we had a simulated fly brain that could help us. >> Yes. And then I need to lock this. How do I lock it? Oh >> yeah. >> Hello overlay locked. >> Lock overlay. Lock overlay. Okay. Now I cannot move it. >> Amazing. >> Yes. Perfect.

16:47 So I I say this because it it seems completely unrelated to Jev, but one of the things you mentioned about Jev is that it responds within 60 milliseconds, right? Which is is really fast. Like that’s enough to self-drive a car, build a robot, etc., etc. >> I think this thing responds in like nano seconds is how it works. Like it’s completely different, but also you have to like you have to invent the carrot or like you have to give it a reward function basically and you have to train it. So, it’s not something that’s ready

17:17 to go. It’s going to take some some actual elbow grease to get to where you want to be, right? >> So, but it feels like we’ve got a few different ways of of doing this at this point in time. So, the overhead of of the flying is going to be astronomical compared to Jev. >> However, if you want something that’s really fast, it looks like you can even get faster. >> Wow. And what kind of hardware it’s running? I mean, this is a simulated brain. So, is it require a big hardware? It doesn’t look like it requires too much. I’m just looking through the

17:48 extension here. You can download the data set, install it using the client. Let me see if I can find the overhead. I thought it ran on CPU. I’m still looking. I thought it was CPU only but we are uh we are checking now and you can download

18:21 it from mail- cns.jl.org. >> It’s pretty interesting. Uh it is in fact required. So it uses WebGL funnily enough. uh two gigabytes of can run it can run in Chrome with hardware acceleration on in the web install Chrome again. >> Well, fair. I think it probably just needs like a two gig uh GPU. I I I will

18:51 test it back because I I am actually kind of interested. Um because I I’ve seen Jad do a lot of stuff that it feels like Fly Brain could do faster because the Doom one I think is basically it’s it’s move left or right or shoot and then if there’s no enemies move forward is is kind of the the algorithm. Left, right, shoot, move forward, left, right or or dodge, I think is what it does. >> It feels like you could train because when when you watch the Fly playing Doom and you watch Jeff playing Doom, I mean the Fly is doing a much better job.

19:22 >> Yeah. I I don’t know if you’ve you’ve compared the two. >> No, no, I had >> And the the organoid acts like a real human, which is like >> actually scary. And I think renting them is is not cheap. >> Wow. >> Wow. Indeed. Okay. I think I’ve mostly run out of news. Oh. Um, did did we talk about DHH yet? Have you seen DH? I can’t even watch the video. of things about him recently, but what was the what’s the gist of the

19:55 >> of it? >> Oh, yeah. So, we we did talk about black pilling a little while ago. So, I feel bad. I hope he didn’t get in like any of I hope the doesn’t [laughter] watch this podcast because we were talking about black pilling and he decided to do it for those in the audience. blackmailing is like instead of picking the red pill or the blue pill, you pick both, which is like a really generally it’s like a it’s a jaded approach I think is what most people will say. It’s like defeist basically. It’s like there is still hope for you as a programmer. I think is generally what we’re trying to say. As a result, you can you can test code, you

20:26 can look at code, you can QA code prevent like there’s a lot of coding that was probably wrote that you didn’t like. Anyway, he he goes into that rant, no problem. And then gets like I I guess really uh stabby [laughter] this dhh and he’s like uh forget like forget coding that’s done. how many people in the audience have coded it in the last week? And then like five people see and then he goes on like I think he swears for a bit. Um [laughter] and then basically says that like the

20:56 future of Rails is not going to be you writing Rails, it’s going to be a machine writing Rails. We’ve got a new release blah blah blah blah. It’s like awesome. I don’t know. People were hurt in a pretty vague way uh [laughter] by basically saying humans don’t have a place in coding anymore. Again, I I was totally I agree. I I haven’t watched the whole keynote. I I should. It’s about five days. >> Well, see, I’ve got this problem there. I mean, he’s he’s Are you the opposite? >> He’s implying that coding in Rails is real coding.

21:27 >> Oh, don’t don’t even get me like as a Ruby fan and like Rails is okay, but like as a Ruby fan and Ruby’s got it own issues. >> Uh but Ruby is supposed to be the >> user friendly programming language. I think that’s what Matt would say that the creator of Ruby. He wanted to be ergonomic basically to make it easy for people to use. >> I don’t know if that’s true, but it it like it probably was easier to use back in the day, but then security and and stuff and it it got pretty bad. Rails on

21:58 top of that these days. I >> Yeah, there there are platforms still running on Rails, I think, famously. Is Spotify one of them? >> Oh, yeah. I >> I think uh well, I guess Bass Camp probably runs on Rails because it’s DHH’s. Um, what else runs on Rails? Shopify maybe. I I can’t remember. Rails apps, but it anyway, we we’ll we’ll ignore his his stuff about

22:28 all the things that DHH is not doing. He also released an operating system that’s called Quattro. So, it’s a fourth version of, and I’m going to say it, Archie. um the uh the Zen approach to Arch and people are saying it’s like the best OS ever and I really still feel like everybody’s missing the point like please make your own Linux. Don’t don’t just use just make your own Linux. I think that’s that’s where we need like that’s the artisal Linux era is here. >> Omari’s got a bunch of plugins and stuff

22:59 that’s pretty cool. I think the the main thing in Omari uh for what was it? It’s got plugins. It’s got bispoke kernel. It’s pretty easy to make a kernel. [laughter] So, sorry. Like I I just I I don’t know. Yeah, it’s Yeah, I I’m I’m not a fan. I will have to try it though because people are telling me now if if I say like you should just make your own Linux and they’re like, “Have you tried Omar?”

23:30 I’m like, “Yeah, I guess I could try it.” >> I’m I’m really not here for it. But yeah. Yeah, that’s I think that’s that’s all we got from the DHA side. I will watch the video, but it was like um >> on the one hand, right? >> He he did. On the one hand, I feel like >> absolutely it’s important. On the other hand, I feel like you could have said this in a much better way, >> which is generally how I feel about DHA. So, I feel like I don’t want to waste my time and get angry at this fool watching the video. Um, I will say one of my big claims to fame back in the day in terms

24:01 of this industry is that I took that, you know, the the photo where he’s given the the crowd the finger. >> I took that photo at Rails Canada. I’m not sure if I’ve [laughter] ever said this a long time ago. Like my most viewed photo on Flickr. I think it had millions of views because people would just embed it in their blogs and stuff like that. And it was copy left so people could use that. I didn’t care. >> Um, but I like I I know of this person. I think that was 2005. You know, it’s it’s been 21 years at this point in time. He hasn’t really changed. >> Yeah. [laughter]

24:32 >> At all. Um, okay. I got two more things. My apologies. I skipped over one from OpenAI dev day. So, they released agents called dots. Did you see the dots release? >> No. >> So, imagine Hermes or Open Writer instead of being like little icons or or like animated companions maybe. Uh, very similar to the rabbit art one or something like that. You know there there’s certainly prior art in this area right now. The absolute best response that I saw in blue sky was uh >> what we need right now is agents represent agent as malware represented

25:04 in the form of a cute avatar. It’s like yeah I’m in here stealing all your data like >> it’s a little on the nose. Uh because it it does feel like that at this point in time if you’re a consumer using the models like like on the free plan or what have you. It does feel like >> it’s really just a cute way to lead your data. And I get why it’ be handy and why that might be okay, but to me it feels like the permissions are are a little bit odd. >> The last one I got here is Open Router Space Bunny Alpha. Have you heard of this one?

25:34 >> No. >> What? >> Okay. It’s It’s really good. I This is maybe me playing with Open Code and Open Router, but not sure where it came from. The rumor is it might be like a new Miniax possibly. >> It’s really good. It’s better than GLM. I think we’re at 5.3 on GLM. It seems to trounce GLM, which is surprising because 5.3 is like actually good. Um, but it looks like this might be minimax. I think is it M3 is in yet?

26:05 Maybe it is. Okay. Yeah, maybe the update to M3 like 3.5 or what have you. >> That is the rumor, but it it does seem to be performing really well. Would suggest you give it a try. I have a feeling this will probably land a lot more soon as well once it’s announced. They typically don’t last too long these like free demo periods. >> I think if I look at this graph, it’s probably got about two days left on its free period. >> Yeah. >> And that’s it. That’s that’s it for my

26:35 news dump. >> They have two bunnies on their uh website. >> Who does? >> Space Bunny Alpha. >> Spacebunnalpha.com. >> They’ve got their own site, too. What’s going on with these these models having their own sites? It’s been unknown. >> Spacebunnymodel.com. >> Spacebunnyalpha.com.

27:05 >> Huh? What is going on? >> Space bunny spelled >> like a space. >> Oh, yeah. Yeah, I found it. >> Okay. >> I just can’t type. Oh, these are AI generators. [clears throat] >> No, >> interesting. These are cute buddies. >> Um, what other uh >> Florida, their attorney general is trying to block OpenAI from doing any

27:36 more model developments. Like, what makes you think Open AI is going to listen to you? >> Interesting. Okay. Yeah. >> Yeah. Like a random state. Is that how laws work? >> No. [laughter] >> Uh, so that’s interesting. I mean, I guess if there’s any Open AI employees in Florida, maybe they’d have a hard time. I don’t know. Um, >> um, well, I’m assuming there are some. >> Yeah, maybe. I mean, there’s some people

28:06 that think Florida’s a nice place to live, >> having grown up in Florida. >> Yeah. I was like, “Do I know anyone who grew up in the Keys?” Yeah. >> Yeah. Well, I mean, I wasn’t didn’t grow up in the the Keys. Kiscane is north of the Keys, even though it’s geologically the same. >> I mean, usually the Keys starts a little bit lower down. >> Gotcha. >> Kibiscane, where I grew up, is is its own thing. >> I mean, it’s you look at the map, it

28:37 looks like it’s Oh, it’s that’s where it starts, but no. Okay. And sorry, my brain went for a little uh walk. [laughter] In Japanese, we might say sample. We’ll we’ll teach you a word while I uh try to collect my thoughts here. >> Um, >> which model do you think would be the best at gator prevention in Florida? Would it be open AI, Jev, or the the fly brain? I I say this is like maybe you don’t need >> gator prevention. Like maybe it’s uh shooting sand pellets, I think, is one

29:08 of the things you can do to animals. and shoot some sand pellets. Like I I don’t know. Maybe Jeb will come up with something better. I feel like OpenAI would architect the solution but wouldn’t know how to run it is maybe where I’m going with this. >> Well, I mean it depends on it depends on the person. I mean, if you go into the Everglades, there’s all the uh um the place you like the showboat kind of a a um places where you pay a little bit money to watch a show and they’ve got the gator guy that like holds the

29:38 gator’s mouth open, in which case you don’t want somebody that the gator to go away. So, I guess it depends on the So, what how for me to avoid being eaten by a gator is the >> I guess I’m thinking of like um that you’ve got in an electric mower, right? What would the electric gator prevention plus AI thing look like? Because I’m thinking like like maybe they don’t need open AI actually. Maybe they just they’re they’re good. They got this covered. >> I I maybe like not anthropic, but Open AI in particular is is also a little bit

30:10 weird to me. It just there’s there’s like nation states making models that [laughter] >> I mean I always like to talk about France as MRO you know uh we’ve got some stuff in here we’ve got Khir in Canada you know are we on the list or or no >> is it the Santos I’m so sorry I should know about US politics it is it is Ronda Santis [laughter] >> yeah um yeah And

30:41 there’s a historically there’s always been a lot of corruption in in Florida government. So, uh uh I wouldn’t be surprised if whoever is going to pay them is the model they’ll use. I don’t know. >> Oh. Oh. Have we stumbled on a a business model? What would the floor man equivalent of a model be? [laughter] >> Some fork of a llama that I’m reselling them. >> Yeah. Oh, that’d be great. >> And it has to be something do something

31:11 stupid to get the level of like Florida man. >> Yeah. It’s kind of like method acting, right? Yeah. You you have to get on that level. Like it’s it is it’s in the it’s um >> the model runs out of a uh a golf cart with two two Mickeys of Jack on [laughter] the side, some random dog or something, and it’s driving around making decisions. >> Totally. >> Ah, sorry. Sorry. I know way too much about the the Florida people stories. [laughter]

31:42 >> Um, so other news, uh, there’s a there’s some company that’s got the Taptic trademark or patent >> that is suing Apple. >> Not not suing, uh, succeeded, won against Apple, have been suing Apple. Uh, this is near and dear to me because like I’ve known about Haptic for a long time. They’re definitely going to appeal. They’ve appealed all of this, but they finally won. was Taptic Games, I think, is what who they were. >> It was like Tapion or Tap Tap

32:12 >> I thought it was Taptic. Taptic Games. >> Yeah. >> Yeah. >> I mean, it seems like a one of these patents that’s an example of why patents are bad or Yeah. [laughter] >> My understanding and this is me attending Sigraph one time. My understanding is this was invented by a Japanese scientist who I’m trying to think when I went to Sigraph. this been like 2011, 2012, like a long time ago. Uh cuz she was I like she read her paper, she gave me a demo. Her English wasn’t that great, but it was like

32:43 really cool tech. And she was like, I I don’t know really where this would >> um where we could use this, but right now on TVs, you know, you’ve got those buttons that you touch and they don’t give you feedback. She was like, maybe there. Love to do it in phones, but the screens are too thick. Like here’s how the motor works. But she was basically giving away her research. She she showed everyone how it worked and she was like and and maybe if you wanted harder you could use a different motor but we couldn’t find one like done right and I was like this is amazing like like when you touch on a screen it feels like you

33:13 touched it. She’s like yeah it’s pretty cool right and >> fast forward Apple added to this product and I was like oh they probably licensed this or or whatever and then fast forward a little bit more and they’re being sued for it and I was like wait a second what [clears throat] happened? So, it sounds to me like Tactic maybe I I don’t know what happens when you go from research to product, but actually did decide to patent it when they did so and Apple >> did not. I I don’t know how that’s even possible or didn’t think they needed to or something or thought that they were prior art, but they definitely weren’t,

33:44 which is odd. >> Yeah. >> Or or it’s a company I mean, I don’t know, a company that >> saw it and just patented it first even though they didn’t >> Yeah. >> have a product, right? cuz I think you can do that too. >> The um uh the Baby Shark song. >> Oh, right. We talked about this. Yeah. Yeah. Yeah. >> Where the uh Pink Fong owned the copyright for it even though they stole it from other people as well.

34:14 So >> yeah. Um, yeah, because it was uh used in some campaign, like a a political campaign, and the campaign got the permission from the people who actually created it. >> Um, and not from Pink Fong, and Pink Fong sued. It’s like, well, no, we got it from these guys who were before you. >> Yeah. Yeah. You actually [laughter] >> who you stole it from. >> Yeah. >> Yes. Sorry. >> Acquired it from. >> Oh, we’re all thieves. Yeah. >> [laughter]

34:46 >> Um, I saw there was also apparently Apple has an Apple Music Hall. >> An Apple Music Hall >> Hall >> like a like a hall for music >> like a where that you can listen to music being played. >> People play actual music still. I’m joking. I’m joking. I’m here for music. >> For the first uh concert they had Elton John sing playing. >> Oh, right on. And it’s the uh I’m not sure how much of the building, but it’s Battery Power Plant. You know, the the

35:17 um one in London. >> Um that’s on >> one of the Pink Floyd album covers. >> Yeah. Only 600 people were present. >> Yeah. >> Must be a small musical. >> Yeah. >> Wow. Uh did find they have a website, appleusic hall.com, which >> today I learned this is incredible. And then >> I I will say just I still own Apple stock, but I’m losing money, folks. So like [laughter] again just like I I do have some bias

35:48 here, but also I’m really good at losing money. >> Yeah. Yeah. Every stock purchase I’ve made is has lost money. >> Yeah. I’m not great at it. I’d be better off just holding stock. >> Yeah. Yeah. [laughter] Keeping under my mattress might be just as productive or more so. >> I I think what uh what do they call that? mattress bank I I think in Japan when you when you cut [laughter] >> an actual term okay >> mattress safe yeah mattress safe >> and then I saw something about Nvidia open shell I don’t know anything about

36:19 it do you know >> ah so yeah um and this is maybe more because I think it’s a bad idea for them to muddy the waters and make their own my understanding is that it’s >> a they say it’s an >> well they say it’s an agent runtime blah blah blah blah blah but what it actually is is kernel secured sandboxes. So >> shrink wrap I think like you know every every operating system has its own API for it. >> The the big aha moment here is that like yes in fact you can use the operating

36:50 systems >> uh native isolation in order to do it. They say host agnostic runtime. Yeah but it uses the APIs to call into it which makes it better >> and as a result I have a feeling like open shell shouldn’t be a thing. Um that’s just me. It’s saying that I think No is better than whatever the heck this is because no’s been around for a while. It’s not a fork of no no again like their their other stuff. It’s very they say it’ll run everywhere, but it feels like it’s

37:21 very Nvidia specific, which being, you know, I’ll leave that there, but basically it’s saying it can connect to like Nvidia Sentry and Bluefield DPU Century, I think, is their AI monitoring product. And Bluefield DPUs are like the physical uh like they can be in a neck or whatever, but like it’s a >> GPIO kind of thing. Like multi-purpose accelerator, so it’s not an MPU or GPU, it’s a >> DPU. Uh so cool. also uh would rather have just the source so like other agents could just

37:52 do this and be secure and and anyway because like they partnered with openclaw this as far as I can tell competes with openclaw it doesn’t it doesn’t add to openclaw so I think that rubbed people the wrong way like I thought y’all were partners you announced all this stuff and then you’re doing this on the other side which very much the same thing >> uh that’s that was my hot take on it I didn’t go any further than that >> but no no no is what I would recommend the judge Check out no.sh. >> Um, and that’s all I Oh, uh, one thing I

38:24 I I wrote down um was that when I wrote down Nimble, uh, when I first saw Nimble, I was thinking, “Oh, wait. Is that the Nimble I know of from before or from another there’s another Nimble?” I mean, it turns out Nimble is a common name for software companies. >> Yeah, because I know a Nimble and I’m wondering if you know the same Nimble. I’m thinking nimbleway.com. >> Absolutely not. >> Okay. And I know that. >> Oh, nimble on lumba. Somehow I missed that. My apologies. >> Nimble on. >> I think you even said that.

38:55 >> I did say that for uh >> having a 38 minute delay in in how your brain’s processing. [laughter] >> I I listen maybe it’s meds for my teeth. I I don’t know. I’m only on antibiotics so I can’t say. >> So Nimble Way uh or that’s our website. Um they do scraping like web scraping. >> Oh. >> Um but the reason I know about them is their I think VP of marketing is ex data

39:25 dog. >> Oh interesting. >> A guy named Rob Denuzo. Do you know Rob at all? >> No. And somehow this company was >> generation data [snorts] dog. Uh, can I can I age myself and give you a blast from the past because because I know of a Nimble, but I think the name was just retired, so it’s kind of like good timing. They did have software, >> but Nimble was a dduplicated all SSD, all flash array. So, this is like, you know, this is within the

39:56 storage sphere is how I know Nimble. But I think the last Nimble array was just retired a month ago because I have a friend who works at HPE and HP bought Nimble and I think they’re I think they’re gone now. Uh but some of the things Nimble did which I thought were interesting was rewriting the file system to write low latency or write to low latency devices because it’s an order of magnitude difference between writing to a hard drive versus writing to an SSD or or

40:27 flash. And as a result, a lot of the quoteunquote contracts or business logic and a flash array actually don’t make sense when you’re using like standard RAID. >> And that was actually pretty cool and sort of changed the entire industry. And as a result, we ended up with a whole bunch of different stuff out there. So I I thought Nimble was amazing back in the day. >> Yeah, >> they solved a lot of problems for me. >> Yeah. >> But yeah, Nimble Way seems to have taken up the moniker now. They’ve got an official plugin for Chat GPD.

40:58 >> Looks cool. >> Yeah. >> Uh I am very much in the market for AI, web scraping, web searching. I think we’ve talked about this a few times, but using Google, even the Brave API, >> you’re going to run out of like Google’s going to start showing you captions within like minutes and then you’ll probably move over to uh I I’ll call it searchng uh is what I use or like the Brave API. The Brave API, you’re going to run into issues as well because it can only batch so many. And I’m not even talking about like a lot of web search. It’s it’s just like actually difficult. And I think we

41:31 had talked about trivia at some point in time because I was using that for a little bit. I think that’s the right name. >> So the one I use occasionally is super data. >> Okay. Not use super data. >> Um >> I think you mentioned this one on my apologies. I think they I feel like I know them because they might have reached out for I don’t know maybe a sponsorship that didn’t actually happen. Um or

42:01 >> but maybe not. Um but then I saw it. It’s like a well I’m I’m buying this. Yeah. So I have a bunch of credits. All right, I’m [laughter] in. >> Um, >> they’re uh it’s from Dump Dumpling Software, which is in EU. Interesting enough. >> Who is >> Dumpling Software? >> Dumpling? >> Yeah. >> What? Dumpling. >> Dumpling AI. Think this is what we found out last time.

42:31 >> Yeah. Dumpling software. >> I would love to eat you. I was thinking when you said dumpling >> I was thinking of >> uh Dump Dumpling the Noodle. >> Ah, gotcha. >> Do you know Dumpling the Noodle? >> I do not know Dumpling the Noodle. >> According to their website, it is the best Chinese restaurant in Seattle and Kirkland, Washington. >> I mean, that’s possible. Sure. Whatever. [laughter] Anybody can say things. I I’m the best

43:02 AI context engineer you’ve ever met. >> Yes. [laughter] Yes. I mean, yeah, best pizza in New York. I mean, all New York pizza is bad, so it’s got to be expert on this channel, on this uh >> Yeah, exactly. [laughter] >> in Seattle on this podcast or or in the Seattle area, sorry, on this podcast. >> Um, oh, shoot. What was I going to say? Super data in fact has a bunch of interesting stuff, but I I just wanted to like pause quickly on the EU fact. Did you hear that Canada? We’re not

43:32 falling out of love with our downstairs neighbors, but we’re we’re just we’re looking around, you know, we’re we’re we’re trying to see if anybody else might accept us into a cool group like NATO. And it turns out the EU is accepting applicants. Uh because Eurovision, I think, if I understand this properly, is going to be hosted by Canada. And this is like our our test. And I’m not saying we got to win, but like listen, if we do okay, >> maybe they’ll let us into the EU is how I understand this problem. Have you heard this this news? I have not heard this, but if that were >> You didn’t know? Oh, man. >> I mean, then why isn’t I mean, Israel’s

44:04 in uh in in the competition, right? But they’re not in the EU. >> I thought they were honorary or something. >> Yes. 2027, we’re in Euro Vision. So, my understanding is we’re already honorary Europeans. we just can’t necessarily [laughter] travel until like if we do well then then maybe a little that’s this is my understanding >> as a former EU citizen >> that’s pretty nice >> now I feel like um I mean the benefits

44:36 are is it multi-use you know canarians there there’s some places I’d like to go in the EU that I’ve not been that aren’t too far >> is the moldi part of EU so let me look this up I’m terrible at geography so this >> [laughter] >> Well, >> what are countries? >> Oh, no. They’re partners actually. >> They’re partners. >> Yeah,

45:08 >> they’re partners. >> What does partner mean? >> As of I don’t know, but as of 1983. So, are they My understanding is you can go to the Mald as an EU citizen very easily. That’s that’s >> the only EU member state not geographically in Europe is Cypress. >> What? Cypress is not in Europe. >> Geolog geographically. I don’t know what that means. >> Yeah. >> Uh it’s it’s generally classified as part of Western Asia. >> Yeah. Interesting.

45:40 European countries EU members include Albania, Andora, Bellarus, Bznia, Herzkabina, Iceland, Kosovo, Likenstein, Mulovo, Monaco. I don’t know why I need to say Likenstein like that. North Macedonia, Norway, Russia, San Marino, Serbia, Switzerland, Ukraine, the United Kingdom, and Vatican City. [laughter] depending >> United Kingdom and Vatican. I knew >> depending on the geographical convention used the transcontinental countries Armenia, Azarbaian, Georgia and Turkey

46:12 may also be included, >> right? Yeah. I always thought it was weird that Turkeykey’s not I I do find it interesting uh that that some countries are and aren’t in the EU. The UK, I think uh they they made their own bed there. I think that’s >> Yeah. >> But Iceland in particular, I keep forgetting, is not part of the >> Yeah. >> And and and sorry, just to go back to Eurovision real quick, like Burke would have won Euro Vision. I mean, >> yeah. I mean, I normally I wouldn’t think of Iceland as being kind of

46:45 European. Then I remember, oh wait, Bork. Yeah. Okay. I could totally see that. >> Yeah, absolutely European. [laughter] >> Yeah. Uh, listen. I I’ I’d love to show you my checkout, but um I’m not wearing it right now. I’m I’m so sorry. >> I I don’t We don’t need to know about purple thongs on this show. That I don’t I don’t need >> I’m pretty sure that’s exactly what the show needs. Uh I did want to talk about Okay, so there’s two things that that I do not have, but I I would like to get.

47:16 The first is the Commodore 64 Ultra. And this is because a teammate purchased one, and I really want one now. Uh they’re they’re not cheap, but there’s a Cyberpunk edition that actually looks awesome. Even just like the keyboard with a really Do you remember the 14 U space bars on the Commodore? It’s got like the actual real long space bar at the bottom. Like underneath all your keys is where the space bar used to be. There was nothing on that ultimate. >> Did I say ultra? I meant my apologies. Yes.

47:47 It’s right in front of me. I just read it wrong. >> Okay. I don’t >> It’s 3.99 USD, which is pretty expensive. >> Yeah. Expensive for something I don’t can’t see it ever wanting. But you would want >> Why? [laughter] >> Okay, hear me out. I understand you can buy a mechanical keyboard on Amazon for about 60 bucks these days. That’s pretty good. Pretty good. I also understand to make a keyboard on my own would probably cost me about $600 just knowing myself. They

48:19 generally cost about that much with key sets and and all the stuff involved. How cool would it be to talk to Claude on a cyberpunk Commodore 64 keyboard >> or do you mean the Commodore is worth 400 bucks? >> Is that is it called this Commodore 77? >> Commodore 77 Cyberpunk. >> Oh, [laughter] I somehow missed that part. That’s terrible. Yeah, I mean I’m here for it because it this looks like the Cloud Code keyboard

48:49 to me. What am I missing? It could run. >> But it’s $377 for the Commodore 77 Cyberpunk. >> Oh, somebody needs to fire this marketing team. That’s awful. Uh, so I don’t have this yet. I I would like to purchase it. I think I did talk about the Superstation One, which has arrived, the FPGA machine. I did finally get it. Uh, I’ve got my first RTL code running on as of yesterday. It’s a USB emulator. It’s terrible, but I I did get one of the FPGA uh units and and again, computers

49:21 that can be other computers is maybe a fascination of mine. And this one to me looks pretty cool. >> So, it has a network jack. >> Yeah, >> it’s pretty cool. >> Yeah, it’s like what are these called? Like cyber decks or something? Have you seen those? Like people brag about their their custom made. Yeah. Yeah. Yeah. This feels like a custom Cyber Deck. >> Doesn’t have a screen, but I could ask her to. >> Cool. >> Uh the the the check out though is the Lululemon heavyweight state joggers. And

49:54 listen, uh what I would show on the stream is a really fat waistband, but it’s like it’s as big as it’s the waistband’s massive. It’s they’re so comfy. They’re oversized. that don’t leave your house in them because no one needs to see you in, you know, gray sweatpants. >> But boy, I I did. Full disclosure, I walked in the rain with them and didn’t get wet because they’re so thick and and like torrential rain. It’s fine. But they’re very thick. Uh >> I think they’re probably $70 US, which is a lot for a pair of pads, but super

50:25 comfy for home. >> And I don’t know. I’m sure for a while when I was still working at uh at Data Dog, >> I was buying >> You didn’t wear any pants? No, I was wearing jeans from uh what was the company called? Seven for Old Mankind or something like that. Or there’s a few different brands. And um I think I had a few pairs of jeans from them and they were usually like 300 bucks. >> Holy moly. >> Yeah, they were good though. >> Like so comfortable. The most

50:56 comfortable. Oh, >> and and and they but they were really thin >> and so they would you need to get them from Nordstrom’s because they would wear out quickly and Nordstrom’s will just replace them and so you can have like add length to the the lifetime. >> Now listen, I I want to I want to laugh at the $300 jeans, but I can’t because I’ve I’ve purchased $300 jeans. Um not from Nordstrom. What is a similar level company in the US? uh like a department

51:27 store that’s pretty expensive. >> I mean, I don’t know if Macy’s still exists. Um I >> think it was Macy’s. Macy’s >> Macy’s might have gone under. I don’t know if they still are on >> I I don’t even think the store is still there, but I was at a >> Yeah, I was at a uh Google whatever the Google conference is. Cloud Next, GCN. I was at a GCN. I was introducing, you know, another speaker. We’re going to go on stage, sit in chairs, and talk to each other. He had brought a full suit and and I definitely did did not [laughter] bring

51:57 a full suit. He’s like, “Oh, well, I’ll be wearing this. What are you wearing?” And I was like, “Well, not that. I mean, I brought jeans, [laughter] a t-shirt.” >> He’s like, “Oh, we should match. I can wear my jeans.” And I was like, “No, you know, like I should actually match your v.” So, I went to Macy’s and I was like, >> “What? Like, it’s tomorrow. What do I do? You got a hemmet.” So, of course, they show sold me some AG jeans, which were like, >> “Yeah.” Well, that’s the one I the other brand I used to buy. AG >> phenomenal feel. They’re they hug everything. Like my butt looks good and

52:27 like as a guy, you know, like wow, my butt looks good. That’s And I bought like five pairs of them after. Like they’re so comfy. Um the problem I’ve got is is similar to what you’re saying. Like my thighs just rub against like I walk a lot. My thighs are rub against each other and it just it it wears apart. So I had one of them the seat like replaced basically. >> The other pair I’ve got still have holes in the crotch area that people just don’t notice. And I again I don’t wear that side too much but yeah I’m here for like the soft comfy jeans after co instead of like the hard

52:58 >> I mean all my jeans beforehand were always bought from Eddie Bower and Eddie Bower is not like the >> are they lifetime though? >> Huh? >> Yeah. >> Well I don’t think >> is Eddie Bower lifime. Oh okay. >> Patagonia. >> I think land Patagonia. >> Yeah. I will say I have an Arcturix jacket that’s got a really nice warranty, but it it leaks water all the time. [laughter] Like it’s such a like I’ve taken it into them multiple times and they’re like, “Oh yeah, just throw it in the dryer.” And I’m like, you know, I’ve done that. It’s still leaking. It’s awful. And I live in a

53:29 very random place as do you. >> You get that uh the the two bottle stuff that you like put a half the bottle or in the with a wash and then do the second bottle and like it re >> I haven’t done that. Oh, I’ll try that. It’s like Gortex repair, I guess. >> Yeah, it’s kind of that kind of thing. >> Okay. >> Um, it didn’t work for me us, but apparently it’s successful for some people. >> Maybe if I had a tent or something. I don’t know. [laughter] >> Um, sorry, I didn’t want to steal your thunder. What are your >> Well, did you happen to So, I’m not

54:02 >> I am not somebody who watches music videos, >> but [clears throat] have you seen video that is like >> I mean there’s one from this person Young Lean or something. Apparently, there was a BMA >> I do know Young Lean >> VMA performance uh yesterday or two days ago >> and it was kind of spectacular. The song is garbage, but the this the the [laughter] choreograph Oh my god. And we I’ve seen so many knockoffs of

54:34 >> Storm. Yeah. Oh, it’s terrible. >> I mean, it’s like the the video is so much better with the volume turned off. >> Um and uh [laughter] uh and the actual video, the sort original video is terrible. I mean, it’s like all the worst stereotypes or like it’s just [laughter] awful. But um but this uh thing that’s on on uh um the VMA uh recording video music awards was it’s like this memerizing movement um which

55:04 just includes like the good part from the actual uh original video the um that I guess did well. Um I don’t know if you watched that. >> I I did not. No. >> But it’s it’s well worth the watch. Um, again, turn the volume down cuz this song is garbage. >> I was googling cuz I was like, I’m pretty sure Young Lean’s from EU. >> Yes, Sweden. >> And they are in fact from Sweden. [laughter] >> Yes. >> I I can only think of this Eurovision. Obviously, there’s nothing else in my

55:35 mind right now. I just I really want like easy flights to Europe. >> And then the other uh checkout uh I have uh moved on on the books. Man, I am going through so many books now because of this thing. >> Now, for the the listeners, Matt is tearing through young adult fiction. Not to call you out, but I’m pretty sure most of the books you tear through are like young adult. >> No. >> Oh, okay. Okay. Okay. >> That is not young adult fiction. I mean, >> I read that when I was a teenager. It’s

56:06 got to be 150 pages long. Yeah. [laughter] >> Well, I was told it was required in English school, so I read it in French. >> Fair enough. Fair enough. And uh what was >> and The Great Gatsby. I mean that’s Yeah. Yeah. That’s that’s that’s for kids too. It’s absolutely for kids. >> It’s not for kids. I mean there’s some >> I I do not >> There are some mature children. >> Yeah. Yeah. >> 16. >> Well, it’s all about wealth. Yeah. >> Um uh but I am now on the adventures of Amina al- Sarafi

56:38 [clears throat] >> by Shannon Chakra Chakraorti. Chakraorti. Yeah. >> Yeah. >> Is she not super famous? >> It It could be. >> Yeah. This like got 90,000 ratings and it’s a 4.25 on Goodreads. I think this is a pretty popular author. Yeah. >> Yeah. Um I think I mentioned I had tried reading that a garbage like terrible book. Um

57:09 it kind of relates to Doom. Uh um Dungeon Crawler Carl or something like that. >> You you did not mention that. Someone recommended that to me and said it’s finally coming out. You should read it. And there’s tons of them, right? [laughter] >> There’s like 10 or 15. I mean, and they are they get a lot of people. I mean, because it’s it’s basically a walkthrough of a video game that this guy is actually playing because the world’s been crushed by aliens. Um, >> it sounds like a bad Ready. Player one,

57:40 choose your own adventure. Is it one? Was it choose your own adventure? >> Like, did you flip the pages? >> Oh, no. You You are not choosing your own adventure. >> Okay. Okay. Okay. >> But that could be cool. Um, but uh this is just I mean watching this guy go through this dungeon and I mean I gave it uh like 20 30% of the book and I was like this is stupid. I’m giving up. [laughter] Yeah, it has a u fervent ravenous like

58:11 the fans are very fanny of dungeon and and I looked at it and I thought this sounds like >> AI choose your own adventure Ready Player One and Ready Player One wasn’t that great of a book >> and it was a pretty bad movie, but I did enjoy it a lot because >> a lot of what they were doing was dealing with multiple video games and giving you like a video game history. one video game we’re doing this thing. >> Yeah. No. And I can’t like I don’t I have no frame of reference for the game that Dungeon Crawler Carl is. Yeah. No.

58:42 Yeah. I’m not here for that. I’ll give it a try. I mean it it looked pretty cheap. I think it was $19 or something. >> Yeah. It’s worth trying, but it’s awful. I mean, it’s it’s a neat concept, but stupid. I I find a lot of times uh science fiction writers suffer the peril of not being able to write um or fantasy authors and it’s like at least if the premise is okay I can deal with the lack of ability but if the premise is bad and the writing is bad then it’s what am I here for just to critique it? >> There was a book that I remember reading

59:14 a bit of and I couldn’t quite finish because it was kind of awful but the premise was so interesting. >> There you go. >> Was it Third Body Problem? Huh? [laughter] >> Three body. >> No, >> the writing was terrible. I had to watch it. >> The one about um uh not rhinos, but oh, hippos. Hippos in the deep south um in and and they got to for cowboys were riding hippos in the deep south.

59:44 >> Okay. >> Uh like 1800s. Um and >> Sure. And and and apparently this was actually a real thing. Like hippos either hippos or rhinos were actually brought >> to the US south around the 18 like 18 mid 1800s I think. >> Amazing >> as a >> I have no idea. >> But they it wasn’t very successful. But this like takes the idea that they were super successful. I think it was it was stupid. It was a bit Yeah. But it was a neat idea. It was funny at first and

1:00:14 then it’s [laughter] like I can’t I can’t finish this. H. That sounds awful. I’m trying to think of a um ah there’s a book I was reading. Oh my my apologies. I’m going to have to look it up on my my actual books. I’m so sorry. Um but I think there are books that are more interesting about rewriting the past than that. And I’m trying to find one here. Uh my apologies.

1:00:45 Oh, it’s not the Hail Mary guy. It’s Oh, shoot. I’m gonna mess this up so bad. I’m so sorry. Uh oh, shoot. Dungeon. Apparently, I pre-ordered Dungeon Carl Carl and it’s in my Apple books to to read. So, I already bought it, which is great. Uh the one I was reading was Alistister Reynolds. He’s got Poseidon’s Children, Blue Remembered Earth, and >> On the Steel Breeze. But basically, it’s a an alter alternate future where Africa because of resources. Well, I guess all futures are alternate futures

1:01:17 >> because they’re futures. >> I think they’re saying from like right now. So, like if we departed right now and said like, okay, >> I don’t know, something happened. China gave Africa a ton of money and then something happened to China. I don’t know, and Africa just went, “Yo, let’s go.” Um, so they they own the space program, etc., etc. I think that series was really really good and and maybe I need to revisit it because it it sort of like changed a lot of the preconceptions I have about space travel and culture and and a lot of that stuff and and would recommend it. It’s actually a really good book. >> The thing I had a hard time with was in

1:01:48 the audio book, the second book had a narrator that did all of the accents of all the different characters. It was actually very distracting. Not that they weren’t good at it. It was more like I was having a hard time understanding the accents, I guess, mostly because I’m just not used to African nation accents. But it was a really really good read and it really changed the way I think about sci-fi. And I I guess that’s where the bar is that I’m setting. And Alistister Reynolds is not sorry Alistister is not a good author. Like they’re they’re serviceable at getting a point across,

1:02:19 but it’s not like okay when you read of Mice of Men it seems plain, but when you think about it, it is not. Yeah. Like the the rain was hitting the rooftop and then falling off the rooftop onto my hat and then off my hat into my cup and it made me think of like that that is a scene like you were there in the scene Alistister Reynolds like and it was raining. [laughter] >> Cool. Moving on. Like we got a setting. Awesome. Sorry, I didn’t mean to go over time, but like I really do feel like Alistister Reynolds is a good awesome >> uh I think I’ve um

1:02:51 uh where’s the list of books I’ve read this finished this um recently? >> Has you gone through like 20 at least, haven’t you? >> Um let’s change this to by year 2026. I’ve gone through 19 books this year. Not bad. >> And that’s really starting >> two months ago. [laughter] >> But some of those I haven’t finished.

1:03:22 Like they were they were >> and some of them >> should we I feel like we should be quizzing you. We we got to set up like a book quiz show where we like just pick random facts from these books you’ve you’ve read just kind of a a what somebody was saying history of the um of the of the loop. There’s Pocket Bear, which is a book that I’ve been reading to my daughter, which was a great story. >> Um, >> you’re not really gonna read all 19 titles, are you? >> Of mice and men. Wombat Waiting, [laughter] which is Oh, wombat waiting. Yeah, Chris Katherine Applegate. She did

1:03:53 um the the best one is the one and only Ivan >> and it’s meant for kids and there’s actually a movie. The book is so it I mean it’ll be it’s a quick read, but it is so great. It is such a great story for everyone. >> The one and only. >> It says it’s a middle-grade novel in verse. >> Yes. And it’s fantastic. >> Infinity Gate was pretty interesting. >> Infinity Gate told us um

1:04:24 >> Yeah, they it takes place in Logos in Nigeria. Uh but she discovers this she creates this device that allows her to jump between different different worlds of of Earth, different versions of Earth, uh different whatever. Um and uh and that was quite fun and spectacular. Um then the one that I think the best of the books I’ve read this year so far is All the Water in the World. >> You did tell us about that one.

1:04:57 >> Totally must. must read. >> All right, you heard it here first. >> Okay, >> books. Read books. >> Read books >> before they’re [laughter] not written by humans anymore. >> And somebody else dh rant. >> One and only Ivan is great. And then there’s like a few companion books because the it’s about Ivan, this gorilla, and his best friend Bob the dog. Um, and so there’s one and only Bob as well, which tells a slightly or

1:05:30 slightly different time period from what the dog thinks. Um, and I think there’s also the one and only either Ruby or Stella because they’re they’re the elephants uh that are in the story. Uh, >> have you seen the movie? It’s not rated very well. >> The the Oh, it was a it was a fun movie. It was a great movie. >> Okay. It’s like Angelina Jolie, Brian Cranston. >> Yeah. Yeah, I mean it definitely makes a diff it’s a little bit different than the book, but I mean it’s still a fun >> It wasn’t really >> I mean it’s Disney.

1:06:01 >> Yeah. No, six out of 10 in Disney. It’s probably pretty good. >> Yeah. >> 72 on Rotten Tomatoes. >> Oh. Oh. Oh. Oh, we got a comment here. The House in the Ceruan Sea in sequel. >> TJ Cloon. Is it? >> Did you know other people read books? >> I apparently [laughter] >> Thank you, Sean. I appreciate it. And apparently somebody showed up because we’re Wow, this is great. >> I I did send out the link to a couple people. So like at least my mom’s on here. >> Oh, yeah. Oh, nice. So, do you know this

1:06:32 Sean McConnell Plus? >> I I do not. Unless that’s my mom’s new nickname online as possible. [laughter] >> Who who I I will admit is trying to write a children’s book and avoid AI and and is very passionate about children’s books. So this is definitely up up your alley. >> Okay. Yeah, I have the house in on here ready to read. So >> there you go. Look at that. And again, I highly recommend of my men. I’m I’m actually a little bit butth hurt that

1:07:03 you didn’t think that was the best book you read [laughter] this whole time. But uh yeah, you know, >> it was I I mean for number two, I’d say Great Gatsby. >> Oh, >> it was so good. Okay. Yeah, we can talk about that one next time. The the morals and obligations of that book not being that yet. >> Fair enough. >> Yeah. Okay. So, I finished. >> Oh, and it was TJ CL. Good. Good call. Look at that. >> Yes. >> Uh, yeah. Let’s find the button. Sorry.

1:07:33 I actually I’m supposed [laughter] to be at the studio. >> Are you supposed to be somewhere? >> Are you blowing off your boss? >> It’s totally fine. No. Sorry. What did you phrasing there? I might have been a little bit not great. Uh >> moving on. >> Yeah, I I’m not sure what I said, but no, that’s >> I’m not playing hookie from work. No. >> Okay. Okay. [laughter] All righty. Bye. >> All right, folks. Okay.