1
00:00:00,540 --> 00:00:05,580
Intro / Outtro: This is GamesItWork.biz, your weekly podcast about gaming, technology, and play.

2
00:00:05,580 --> 00:00:08,980
Intro / Outtro: Your hosts are Michael Martine, Andy Piper, and Michael Rowe.

3
00:00:08,980 --> 00:00:12,680
Intro / Outtro: The thoughts and opinions on this podcast are those of the hosts and guests alone,

4
00:00:12,680 --> 00:00:18,200
Intro / Outtro: and are not the opinions of any organization which they have been, are, or may be, affiliated with.

5
00:00:18,200 --> 00:00:23,440
Intro / Outtro: This is Episode 567, Infinite Pong World Model.

6
00:00:30,000 --> 00:00:34,320
Andy Piper: Hello, welcome back to gamesatwork.biz. My name is Andy Piper.

7
00:00:34,320 --> 00:00:39,680
Andy Piper: I am joined by my friend and co-host for this week, Michael Martine.

8
00:00:39,680 --> 00:00:44,820
Andy Piper: We're missing Michael Rowe again this week, but I'm sure we'll have him back again before too long.

9
00:00:44,820 --> 00:00:49,020
Andy Piper: As you may know, if you're a listener, then we are your weekly technology podcast.

10
00:00:49,020 --> 00:00:53,480
Andy Piper: We may not be your only one, but we can be your only one, should you need it.

11
00:00:53,480 --> 00:00:54,560
Andy Piper: Should you need it?

12
00:00:54,560 --> 00:00:55,460
Andy Piper: We should be, probably.

13
00:00:56,180 --> 00:01:00,220
Andy Piper: If you're a new listener, you just discovered us, you may have come across our link somewhere,

14
00:01:00,220 --> 00:01:04,880
Andy Piper: shared in the Fediverse, in email, your friends have told you,

15
00:01:04,880 --> 00:01:08,900
Andy Piper: you heard one of us speaking at an event, and we said, listen to our podcast, wherever.

16
00:01:08,900 --> 00:01:10,000
Michael Martine: Skywriting.

17
00:01:10,000 --> 00:01:10,300
Andy Piper: Yeah.

18
00:01:10,300 --> 00:01:10,700
Andy Piper: You know.

19
00:01:10,700 --> 00:01:13,080
Andy Piper: Welcome. We're delighted to have you.

20
00:01:13,080 --> 00:01:15,680
Andy Piper: We talk about fun stuff on the internet every week.

21
00:01:15,680 --> 00:01:16,660
Andy Piper: Michael, how are you?

22
00:01:16,660 --> 00:01:20,540
Michael Martine: I'm doing very well, and excited to spend some time with you here again too.

23
00:01:20,540 --> 00:01:26,720
Michael Martine: So let's get to it, because we've got an interesting thing that we've been talking about here recently,

24
00:01:26,720 --> 00:01:31,320
Michael Martine: about e-ink readers and doing some interesting things with the firmware.

25
00:01:31,320 --> 00:01:34,760
Michael Martine: And you found something here about playing Pokemon.

26
00:01:34,760 --> 00:01:35,200
Andy Piper: Yes.

27
00:01:35,200 --> 00:01:41,920
Andy Piper: So I talked last week about, we both talked last week about these little handheld palm-sized e-ink readers.

28
00:01:41,920 --> 00:01:46,740
Andy Piper: There's a brand called XteInk, which is doing them and a couple of others as well.

29
00:01:47,320 --> 00:01:51,440
Andy Piper: And the customizable firmware, so you can absolutely use them for reading books or articles,

30
00:01:51,440 --> 00:01:52,660
Andy Piper: which is what I use mine for.

31
00:01:52,660 --> 00:01:58,600
Andy Piper: But you can also, as we discovered last week, you can also put on some interactive fiction, for example.

32
00:01:58,600 --> 00:02:03,680
Andy Piper: But something I found is this kind of Pokemon game.

33
00:02:03,680 --> 00:02:06,360
Andy Piper: Yeah, it's more like, I guess, the card game.

34
00:02:06,360 --> 00:02:10,560
Andy Piper: And you can, yeah, just do some little things.

35
00:02:10,560 --> 00:02:14,900
Andy Piper: It's not a sort of super active game, because it's obviously e-ink.

36
00:02:14,900 --> 00:02:23,520
Andy Piper: But it's really quite interesting the way that people are adding these kind of gamification to reading or other devices.

37
00:02:23,520 --> 00:02:24,060
UNKNOWN: Yeah.

38
00:02:24,060 --> 00:02:29,860
Michael Martine: And I like how it's structured so that you're reading instead of walking around.

39
00:02:29,860 --> 00:02:35,220
Michael Martine: And that may cause a wild Pokemon to appear that you can potentially go catch.

40
00:02:35,220 --> 00:02:36,500
Michael Martine: So that's kind of fun, right?

41
00:02:36,500 --> 00:02:36,740
Michael Martine: Yeah.

42
00:02:36,740 --> 00:02:38,440
Michael Martine: Encouraging people to read, to catch Pokemon.

43
00:02:38,580 --> 00:02:38,940
Michael Martine: I like it.

44
00:02:38,940 --> 00:02:44,680
Andy Piper: This is a shame for me, because I have an X4 reader, and this one is for the XTE and X3.

45
00:02:44,680 --> 00:02:49,460
Andy Piper: I'm sure it potentially shouldn't be too difficult, because it's mostly a screen size difference.

46
00:02:49,460 --> 00:02:50,900
Andy Piper: But I'll have to have a look.

47
00:02:51,140 --> 00:02:58,280
Michael Martine: So now we're going to have to talk a little bit about Pong, you know, one of the very early, early games, you know, even before e-ink readers were around.

48
00:02:58,280 --> 00:03:06,760
Michael Martine: And this example on Steam called Pong Quest has a lot more to it than just the paddles going back and forth, doesn't it?

49
00:03:07,000 --> 00:03:13,420
Andy Piper: I was going through some old emails where I used to send myself links to things to look up.

50
00:03:13,420 --> 00:03:17,200
Andy Piper: And I found some old, old emails from 2020.

51
00:03:17,200 --> 00:03:19,980
Andy Piper: So pre-pandemic or mid-pandemic time.

52
00:03:19,980 --> 00:03:27,200
Andy Piper: And at the time, there was this announcement that Atari was going to be bringing out this RPG version of Pong.

53
00:03:27,280 --> 00:03:29,060
Andy Piper: And I thought, that's crazy.

54
00:03:29,060 --> 00:03:30,480
Andy Piper: Did they ever do it?

55
00:03:30,480 --> 00:03:30,960
Andy Piper: Because I didn't.

56
00:03:30,960 --> 00:03:35,160
Andy Piper: And then I looked up to check whether we'd talked about it on the show before, and I can't find any evidence of it.

57
00:03:35,160 --> 00:03:39,440
Andy Piper: Atari owns the Pong license, the franchise.

58
00:03:40,660 --> 00:03:45,840
Andy Piper: And so they've made Pong into kind of this roguelike dungeon crawler.

59
00:03:45,840 --> 00:03:48,600
Andy Piper: So I had to download it on the Steam Deck and give this a try.

60
00:03:48,600 --> 00:03:57,360
Andy Piper: It is Pong, but you go through a series of kind of, your little paddle goes through a series of dungeons to find people to fight.

61
00:03:57,360 --> 00:04:05,100
Andy Piper: And then when you fight and play Pong, there's little tokens to give you, you know, additional powers or whatever.

62
00:04:05,100 --> 00:04:16,860
Andy Piper: And it gets into almost the sort of Arkanoid breakout kind of situation where you can build one of the tokens you get lets you have like a wall of bricks in front of you so that you can defeat the other person more easily.

63
00:04:16,860 --> 00:04:18,200
Andy Piper: It's crazy.

64
00:04:18,200 --> 00:04:22,000
Andy Piper: It's a whole mashup of genres going on here.

65
00:04:22,000 --> 00:04:25,220
Andy Piper: But yeah, those crazy people went and did it.

66
00:04:25,220 --> 00:04:25,700
Michael Martine: Wow.

67
00:04:25,700 --> 00:04:27,940
Michael Martine: And this is from Atari, isn't it?

68
00:04:27,940 --> 00:04:28,500
Andy Piper: Yeah.

69
00:04:28,500 --> 00:04:29,320
Michael Martine: Amazing.

70
00:04:29,320 --> 00:04:29,500
Michael Martine: Yeah.

71
00:04:29,500 --> 00:04:34,800
Michael Martine: Well, I mean, that's a great way to repurpose your intellectual property.

72
00:04:35,100 --> 00:04:38,300
Michael Martine: And continue to bring it back, you know, in a new and a different way.

73
00:04:38,300 --> 00:04:40,260
Andy Piper: Of course, you can modify your paddle.

74
00:04:40,260 --> 00:04:42,300
Andy Piper: You're not just a little white rectangle.

75
00:04:42,300 --> 00:04:47,840
Andy Piper: As you go through the game, you also collect outfits and headgear and accessories and...

76
00:04:47,840 --> 00:04:50,640
Michael Martine: All the things that you might need if you were playing Pong.

77
00:04:50,640 --> 00:04:50,920
Andy Piper: Definitely.

78
00:04:50,920 --> 00:04:53,480
Michael Martine: So I found something fun here.

79
00:04:53,480 --> 00:04:58,980
Michael Martine: And this is not the first time I've seen it, but there's a problem sometimes people have.

80
00:04:58,980 --> 00:05:03,780
Michael Martine: And that is when you're starting a game and you have to roll to see who gets to go first, right?

81
00:05:03,820 --> 00:05:04,840
Michael Martine: Who's the first person?

82
00:05:04,840 --> 00:05:09,680
Michael Martine: Every time you roll a regular die six, you know, you might get a tie, right?

83
00:05:09,680 --> 00:05:10,480
Michael Martine: I might roll a three.

84
00:05:10,480 --> 00:05:11,560
Michael Martine: You might roll a three, Andy.

85
00:05:11,560 --> 00:05:13,440
Michael Martine: And that's super frustrating, isn't it?

86
00:05:13,440 --> 00:05:14,760
Michael Martine: Because then we have to roll again.

87
00:05:14,760 --> 00:05:19,400
Michael Martine: And, you know, wouldn't it be nice if we could just solve that problem once and for all?

88
00:05:19,400 --> 00:05:26,300
Michael Martine: Well, a group of mathematicians talking about this have created what they call go first dice.

89
00:05:27,100 --> 00:05:32,120
Michael Martine: And what it is is several die that you can all roll at the same time.

90
00:05:32,120 --> 00:05:38,520
Michael Martine: And the way those dice are set is that there will always be a winner on the first roll for all these dice.

91
00:05:38,520 --> 00:05:40,360
Michael Martine: So there you go.

92
00:05:40,360 --> 00:05:41,700
Michael Martine: Mathematically, okay.

93
00:05:41,940 --> 00:05:46,560
Michael Martine: And I don't remember how many sides these die have, but there's a lot.

94
00:05:46,560 --> 00:05:55,960
Michael Martine: And the first time I saw it, this team had made, like, you know, not small models, like for your tabletop, but like really big ones that you could even sit on top of.

95
00:05:55,960 --> 00:05:56,940
Michael Martine: I mean, they were that large.

96
00:05:56,940 --> 00:05:58,260
Michael Martine: So fun stuff.

97
00:05:58,920 --> 00:06:06,820
Andy Piper: So the idea, I think, is also that they are, well, first of all, it does describe them as a theoretical set of dice, but I presume they've actually made them now.

98
00:06:06,820 --> 00:06:07,240
Michael Martine: Yes.

99
00:06:07,240 --> 00:06:11,320
Andy Piper: And the dice don't all have the same numbers on the faces.

100
00:06:11,320 --> 00:06:20,980
Andy Piper: So across the set, I think, they've distributed enough numbers across the faces that you end up with unique winners all the time.

101
00:06:21,120 --> 00:06:28,460
Andy Piper: As far as I can tell from the image, because I can't read, I haven't read the whole article because it's on the new scientist and it's paywalled, which is just rude.

102
00:06:28,460 --> 00:06:30,860
Andy Piper: I would like to like to know more about it.

103
00:06:30,860 --> 00:06:32,200
Andy Piper: It is a fascinating idea.

104
00:06:32,200 --> 00:06:32,660
Andy Piper: Yeah.

105
00:06:32,660 --> 00:06:33,800
Michael Martine: Very good.

106
00:06:33,800 --> 00:06:37,580
Andy Piper: We like, we like novel applications of science and mathematics.

107
00:06:37,580 --> 00:06:38,620
Michael Martine: And dice.

108
00:06:38,620 --> 00:06:38,960
Andy Piper: And dice.

109
00:06:38,960 --> 00:06:47,800
Michael Martine: So thinking a little bit further into a concept that I've been kicking around in my head here for a while, which is personalized software.

110
00:06:48,520 --> 00:06:53,000
Michael Martine: Software that can be done for a audience of one.

111
00:06:53,000 --> 00:06:53,180
Andy Piper: Yeah.

112
00:06:53,180 --> 00:06:54,160
Michael Martine: Even for yourself.

113
00:06:54,160 --> 00:06:57,120
Michael Martine: This article from Ars Technica caught my eye.

114
00:06:57,120 --> 00:07:05,500
Michael Martine: And that was because in the Meta universe, there is a new set of capability that meta acquired.

115
00:07:05,500 --> 00:07:08,080
Michael Martine: They were a separate company along the lines before.

116
00:07:08,080 --> 00:07:18,040
Michael Martine: A pocket AI to be able to create what they call gizmos, which are small games that you can describe in very much a vibe coding kind of fashion.

117
00:07:18,320 --> 00:07:18,440
Michael Martine: Right.

118
00:07:18,440 --> 00:07:26,700
Michael Martine: So I want a game where I have a Pong paddle that can run around in a roguelike environment and, you know, go do some fun stuff with a fun.

119
00:07:26,700 --> 00:07:27,380
Andy Piper: I should try this.

120
00:07:27,380 --> 00:07:34,320
Andy Piper: And try programming Pong RPG as using using this pocket AI.

121
00:07:34,320 --> 00:07:34,660
Michael Martine: Yeah.

122
00:07:34,660 --> 00:07:34,980
Michael Martine: Yeah.

123
00:07:35,100 --> 00:07:38,380
Michael Martine: So the notion here is quite simple.

124
00:07:38,380 --> 00:07:44,640
Michael Martine: The intriguing thing about it is that because this is a walled garden kind of approach, there's no code here.

125
00:07:44,640 --> 00:07:59,000
Michael Martine: There's no way to port what you have created and your interactions with the large language model to create this software, which is good for you and maybe is shareable amongst others, which is the design point to keep you inside the meta environment.

126
00:07:59,480 --> 00:08:02,460
Michael Martine: And let it be an entertainment thing for others.

127
00:08:02,460 --> 00:08:07,280
Michael Martine: But you can't necessarily take it with you or use it in other spots and other places.

128
00:08:07,860 --> 00:08:11,380
Michael Martine: So that was an interesting thing for me.

129
00:08:11,380 --> 00:08:18,840
Michael Martine: So it's writing code, but that code, those gizmos, as they call them, only operate within the pocket AI environment.

130
00:08:19,080 --> 00:08:20,140
Andy Piper: So it reminds me of two things.

131
00:08:20,140 --> 00:08:30,780
Andy Piper: It reminds me, first of all, of Facebook apps back in the day when Facebook was very early on, where they were encouraging people to build things that connected into and ran inside Facebook.

132
00:08:30,780 --> 00:08:32,360
Andy Piper: And, of course, they moved away from that in the end.

133
00:08:32,360 --> 00:08:34,100
Andy Piper: But, again, they were walled gardens.

134
00:08:34,100 --> 00:08:37,900
Andy Piper: You know, the whole purpose of them was to get more people to have a Facebook account.

135
00:08:38,960 --> 00:08:40,700
Andy Piper: You needed the Facebook account to do that.

136
00:08:40,700 --> 00:08:47,000
Andy Piper: And, ultimately, that, I suppose, ended up becoming their whole ecosystem of, OK, we're going to buy Instagram.

137
00:08:47,000 --> 00:08:52,820
Andy Piper: You have to have a Facebook account or a Meta account to use Instagram and the VR stuff and so on as well.

138
00:08:52,820 --> 00:08:56,240
Andy Piper: But it's interesting to me that they acquired Gizmo, which became Pocket.

139
00:08:56,240 --> 00:09:01,500
Andy Piper: It turns out it's only launched in the US so far, I think, but reading the article, so I may not be able to test it.

140
00:09:01,560 --> 00:09:07,100
Andy Piper: The idea that it is built deeply into the platform, you can't port them outside, is a challenge.

141
00:09:07,100 --> 00:09:25,540
Andy Piper: The other thing it reminds me of, though, a little bit, is the era of Flash games and other, you know, run times that required you to use sort of that platform or that wrapper to run the software.

142
00:09:26,100 --> 00:09:37,440
Andy Piper: So, famously, you know, Steve Jobs wanted to kill Flash and have everything run on the web at a certain point in time before then pivoting towards everybody should be writing iOS apps.

143
00:09:37,440 --> 00:09:46,860
Andy Piper: But the notion that there are now all of these old Flash games that, because the Flash runtime is no longer a thing, you can't play anymore.

144
00:09:46,860 --> 00:09:48,540
Andy Piper: It's similar to this, right?

145
00:09:48,540 --> 00:09:53,960
Andy Piper: You can't do anything with these apps outside of the Meta and Pocket platform.

146
00:09:53,960 --> 00:10:02,740
Andy Piper: If they never open up the platform and the runtime to let you port that thing elsewhere, then, yeah, it's either throwaway or just deeply locked in.

147
00:10:02,740 --> 00:10:10,800
Michael Martine: So, like I said, I'm really intrigued right now about this whole idea of software for a universe of one, a user of one.

148
00:10:10,800 --> 00:10:17,780
Michael Martine: And that was something I did actually myself last weekend because I was frustrated by something.

149
00:10:17,860 --> 00:10:20,520
Michael Martine: And I was like, I am going to solve this problem for me.

150
00:10:20,520 --> 00:10:21,460
Michael Martine: And you know what?

151
00:10:21,460 --> 00:10:30,560
Michael Martine: I was able to do so through the Vibe coding process and create the Swift code that could run via test flight on my iPhone.

152
00:10:30,560 --> 00:10:35,620
Michael Martine: And it is solving the problem that I had, which is like, this is pretty darn cool.

153
00:10:35,620 --> 00:10:42,060
Michael Martine: Now, it also means that there's going to be a lot of code in a lot of places that is going to be a little bit more difficult to maintain over time.

154
00:10:42,460 --> 00:10:45,480
Michael Martine: But if you're willing to do it, it's a possibility.

155
00:10:45,480 --> 00:10:46,040
Michael Martine: Yeah.

156
00:10:46,040 --> 00:10:46,640
Michael Martine: Yeah.

157
00:10:46,640 --> 00:11:01,460
Michael Martine: So, speaking of running code and having some walled gardens that are now world-sized, I read today a little bit about this new Solaris capability from Runway.

158
00:11:01,660 --> 00:11:20,720
Michael Martine: And what they're calling an interface world model is instead of doing an LLM, just sort of writing code, it's generating the application and the video visualization directly.

159
00:11:20,720 --> 00:11:22,220
Michael Martine: There's not code.

160
00:11:22,220 --> 00:11:23,560
Michael Martine: It is images.

161
00:11:23,560 --> 00:11:27,660
Michael Martine: And the way they call it is, quote, unquote, the image is the application.

162
00:11:28,040 --> 00:11:39,340
Michael Martine: So, this little video snippet that we'll include in the show notes below shows examples of a 3D environment where it's an empty room and you can start to populate the room with furniture.

163
00:11:39,340 --> 00:11:43,800
Michael Martine: And you want to move the lamp from one corner to the other and you're moving things around.

164
00:11:44,020 --> 00:11:56,000
Michael Martine: And that is the environment of a world-building set of capabilities, which reminded me an awful lot of the NVIDIA OmniWorld kind of environment, too.

165
00:11:56,000 --> 00:12:03,540
Michael Martine: We're collecting all that data about buildings and trees and topography and all kinds of other stuff could be ingested in.

166
00:12:03,540 --> 00:12:08,240
Michael Martine: And then you could change it to be something other than what it is.

167
00:12:08,380 --> 00:12:10,740
Michael Martine: And I found this incredibly interesting.

168
00:12:10,740 --> 00:12:13,260
Andy Piper: So, I found the video intriguing.

169
00:12:13,260 --> 00:12:20,400
Andy Piper: I didn't exactly understand what it was trying to sell me or what I could do with it or how I would engage with it.

170
00:12:20,400 --> 00:12:26,700
Andy Piper: Looking at their company website, it's pitched very much, I think, at creatives.

171
00:12:26,700 --> 00:12:40,540
Andy Piper: It's pitched in the media creative space around building video content, which you can animate, modify, do photorealistic simulation, things like that.

172
00:12:40,540 --> 00:12:48,340
Andy Piper: I think the thing that we both agreed on is their choice of language and terminology around this general world model.

173
00:12:48,660 --> 00:12:51,520
Andy Piper: Rather than talking about the language, they're talking about the world.

174
00:12:51,520 --> 00:13:04,220
Andy Piper: So, really modeling entities, objects, real physical things that you can then, at the snap of the fingers, change the color, the texture, the layout, the lighting, and all those kinds of things.

175
00:13:04,480 --> 00:13:14,460
Andy Piper: And they've got some really fantastic, I'll be honest, videos and clips embedded in their homepage demonstrating lots of creative concepts.

176
00:13:14,460 --> 00:13:25,260
Andy Piper: You know, a man, a small child dressed as a schoolboy walks up to two more school boys, turns the hands on his wristwatch and grows up into an adult.

177
00:13:25,260 --> 00:13:29,060
Andy Piper: Photo realistically in black and white as a film clip.

178
00:13:29,060 --> 00:13:30,220
Andy Piper: Looks fantastic.

179
00:13:30,220 --> 00:13:31,240
Andy Piper: How does it work?

180
00:13:31,240 --> 00:13:32,440
Andy Piper: What I do with that?

181
00:13:32,440 --> 00:13:35,140
Andy Piper: What is the offer?

182
00:13:35,140 --> 00:13:38,440
Andy Piper: So, I hadn't come across them.

183
00:13:38,440 --> 00:13:42,660
Andy Piper: They've evidently been around for a while and clicking through and digging around.

184
00:13:42,660 --> 00:13:48,160
Andy Piper: It seems like they had something to do with stable diffusion at some stage, the folks that set this up.

185
00:13:48,160 --> 00:13:52,760
Andy Piper: So, I'm going to do a bit of digging and learning because I'm curious what this is.

186
00:13:52,760 --> 00:13:58,480
Andy Piper: But I'm certainly intrigued by the notion of this world model that they're pitching.

187
00:13:58,480 --> 00:13:59,680
Michael Martine: Yeah.

188
00:13:59,680 --> 00:14:02,140
Michael Martine: And I need to do some more reading on this as well.

189
00:14:02,140 --> 00:14:09,340
Michael Martine: It was something that came across my inbox earlier today from the Not Boring email that I subscribe to.

190
00:14:09,340 --> 00:14:11,900
Michael Martine: And there's more here to this.

191
00:14:11,900 --> 00:14:14,160
Michael Martine: So, a little bit of a starting point for this.

192
00:14:14,160 --> 00:14:17,280
Michael Martine: And I'm sure we're going to come up with even more as we go along.

193
00:14:17,480 --> 00:14:22,200
Andy Piper: But as Mike Elgan posts on Mastodon, you know, all of this AI stuff is all very well.

194
00:14:24,200 --> 00:14:34,740
Andy Piper: It links to a computer world article that doesn't automatically make companies more productive just by starting to use it in a big organization.

195
00:14:34,960 --> 00:14:36,360
Andy Piper: We've heard this time and time again now.

196
00:14:36,360 --> 00:14:45,460
Andy Piper: What actually happens quite often as a tool, maybe it helps some people get through their day more quickly, fulfill a particular task more quickly sometimes.

197
00:14:46,000 --> 00:14:51,660
Andy Piper: But of course, one of the reasons that that falls down is that you then still need to do a lot of tidying up.

198
00:14:51,660 --> 00:14:54,200
Andy Piper: You need to really direct the models.

199
00:14:54,200 --> 00:15:07,100
Andy Piper: There's a lot of babysitting around what agents are doing, whether or not they're following the instructions as defined, because they are not necessarily going to do that, because that's not how the technology works.

200
00:15:07,260 --> 00:15:23,520
Andy Piper: There's this whole article here talking about various reasons why AI doesn't necessarily make everybody more productive or big companies more productive, especially when you aggregate that up to hundreds of workers all fighting their own AI problems.

201
00:15:23,520 --> 00:15:37,740
Michael Martine: You know, I would say that this week, at least for me, I am an anomaly from this article because I found new uses that helped me do things that otherwise would have been incredibly time consuming for me to do.

202
00:15:37,740 --> 00:15:39,840
Michael Martine: So I'll give you one example.

203
00:15:39,840 --> 00:15:52,800
Michael Martine: While we do this podcast in an unscripted form, and you can probably tell it's unscripted, my friends, I do from time to time put a video blog out in my company about things that have happened in the past week.

204
00:15:52,900 --> 00:16:00,980
Michael Martine: And in order to do that sort of thing, I like to have a sort of a newsy kind of tone that's punchy, that's interesting, that's good.

205
00:16:00,980 --> 00:16:09,100
Michael Martine: And it requires me to take an input, look at the inputs, refactor it in my style, in my tone, and then record.

206
00:16:09,100 --> 00:16:17,160
Michael Martine: And that first step there ordinarily will take me somewhere between 20 and 30 minutes or so to have that done.

207
00:16:17,160 --> 00:16:22,780
Michael Martine: And then I have it in a document that I can now have in front of me, have the screen in front, I could hit record, and off we go.

208
00:16:23,480 --> 00:16:28,240
Michael Martine: So I've created a structure where I've said, here's some examples of how I've done it before.

209
00:16:28,240 --> 00:16:31,260
Michael Martine: Here's the source, and this is what I created in writing.

210
00:16:31,260 --> 00:16:32,800
Michael Martine: Here's a new one.

211
00:16:32,800 --> 00:16:33,940
Michael Martine: Please do the same thing.

212
00:16:33,940 --> 00:16:39,860
Michael Martine: It was amazing to find that this week that definitely did save me 30 minutes to go and create the script.

213
00:16:39,860 --> 00:16:41,760
Michael Martine: So I was able to record it that much faster.

214
00:16:41,980 --> 00:16:44,020
Michael Martine: I got it one take, so I was happy about that.

215
00:16:44,020 --> 00:16:45,360
Michael Martine: And off I went.

216
00:16:45,360 --> 00:16:50,540
Michael Martine: And on top of that, I then said, this seems like it would make a good skill.

217
00:16:50,540 --> 00:16:57,480
Michael Martine: So please create a skill markdown file for me so that the next time I do it, I'm going to have this ready to rock and roll.

218
00:16:57,560 --> 00:17:00,760
Michael Martine: And that was exactly what I did.

219
00:17:00,760 --> 00:17:10,200
Michael Martine: And I've been playing even more with skills in that fashion to be able to then string them together, to remix them in different ways.

220
00:17:10,460 --> 00:17:13,900
Michael Martine: And like I said, I'm finding this to be absolutely fantastic.

221
00:17:14,040 --> 00:17:20,340
Andy Piper: Well, this is actually Mike Elgan's article in Computer World that we got the link to.

222
00:17:20,340 --> 00:17:27,940
Andy Piper: And as he says at the end, as we say on all of our episode blog posts, this article is 100% human written by the author.

223
00:17:27,940 --> 00:17:32,540
Andy Piper: Although he also then points out that some grammar checking and so on was performed using AI tools.

224
00:17:32,540 --> 00:17:38,660
Andy Piper: The point that he also makes here is that there's the micro or individual level, which you've just described, your experience.

225
00:17:38,980 --> 00:17:44,580
Andy Piper: And then there's the macro level of everybody at aggregate scale doing all the same sort of things.

226
00:17:44,580 --> 00:17:47,380
Andy Piper: And then not everybody is having that same experience.

227
00:17:47,380 --> 00:17:47,760
Andy Piper: Absolutely.

228
00:17:47,760 --> 00:17:48,100
Andy Piper: Interesting.

229
00:17:48,100 --> 00:17:48,620
Andy Piper: Interesting.

230
00:17:48,620 --> 00:17:49,200
Andy Piper: Very good.

231
00:17:49,200 --> 00:17:54,240
Andy Piper: We are still we're still not introducing it directly to our workflow for creating our podcast.

232
00:17:54,240 --> 00:18:04,340
Andy Piper: But eventually, eventually you'll probably write a skill that replaces me and just says, could you just get a British guy, get a British accent agent to correct us?

233
00:18:04,340 --> 00:18:06,220
Michael Martine: So funny you should say that, right?

234
00:18:06,280 --> 00:18:10,060
Michael Martine: Because couldn't I apply the same idea, you know, to doing the show notes?

235
00:18:10,060 --> 00:18:12,520
Michael Martine: And the answer to that is I could, right?

236
00:18:12,520 --> 00:18:17,880
Michael Martine: Because we do generate a transcript and could we ingest the transcript and kind of structure show notes in that way?

237
00:18:17,880 --> 00:18:27,400
Michael Martine: And I don't do that because it is human created the show notes because I want to be experiencing, you know, what did we talk about?

238
00:18:27,400 --> 00:18:30,520
Michael Martine: And how do I refactor in a way that makes the most sense to me?

239
00:18:30,580 --> 00:18:32,600
Michael Martine: So I have not gone down that path.

240
00:18:32,600 --> 00:18:38,060
Michael Martine: And we do have our closing comment, as you rightly point out, that they were lovingly created by hand.

241
00:18:38,060 --> 00:18:40,240
Michael Martine: And that's still the process, at least for now.

242
00:18:40,240 --> 00:18:42,120
Andy Piper: Well, there's one more thing in this section.

243
00:18:42,120 --> 00:18:45,120
Andy Piper: And then we've got a couple of fun things to talk about to close out.

244
00:18:45,360 --> 00:18:51,780
Andy Piper: And that is the there was a blog post on the Linux kernel maintainers in the last week entitled Creepy Crawlies.

245
00:18:51,780 --> 00:18:54,000
Andy Piper: And we've got a link in the show notes here again.

246
00:18:54,000 --> 00:19:07,260
Andy Piper: And it's demonstrating the blunt force that a lot of these tools continue to use whilst companies are claiming that they're building this super great intelligence for us.

247
00:19:07,260 --> 00:19:19,980
Andy Piper: Actually, what I believe they've done is they've given a lot of very basic tools to a very large number of compute engines and let them all mine at the same time.

248
00:19:19,980 --> 00:19:35,740
Andy Piper: And so what's happening on the Linux kernel server, for example, is that all of these robots are showing up at the door and constantly, constantly requesting HTML rendered diffs of code in as expensive a manner as you possibly could.

249
00:19:36,220 --> 00:19:54,220
Andy Piper: To the extent that, you know, the vast proportion of compute that's used by these servers on the Linux kernel legitimate Git server, the place where all the source code is stored, is servicing these utterly dumb requests for historical source comparisons.

250
00:19:54,220 --> 00:19:56,460
Andy Piper: And he's got the charts.

251
00:19:56,460 --> 00:19:57,920
Andy Piper: He brings the charts to back it all up.

252
00:19:57,920 --> 00:20:07,360
Andy Piper: It really does make you realize that how limited the intelligence is in a lot of these things.

253
00:20:07,360 --> 00:20:13,420
Andy Piper: As I said, I think really what they are is, I think you used in our pre-show conversation, the term bulldozer.

254
00:20:13,420 --> 00:20:19,380
Andy Piper: There's a load of idiots with bulldozers showing up to try and bulldoze the same property at the same time.

255
00:20:19,380 --> 00:20:26,440
Michael Martine: You know, I was interested as we read about this, I was looking at Anubis, which was one of the ways of kind of protecting against this.

256
00:20:26,440 --> 00:20:33,700
Michael Martine: And even when I looked at the website, it had to figure out, am I a person or am I a bot scraping too?

257
00:20:33,700 --> 00:20:34,320
Michael Martine: Yeah.

258
00:20:34,320 --> 00:20:35,260
Michael Martine: So that was quite cool.

259
00:20:35,540 --> 00:20:41,460
Andy Piper: So Anubis is a tool for trying to prevent this, but really all it does is slow them down.

260
00:20:41,460 --> 00:20:46,860
Andy Piper: They just continue because they don't have willpower and intelligence to give up.

261
00:20:46,860 --> 00:20:52,340
Andy Piper: Their willpower is entirely, as it exists, is entirely driven towards succeeding in their task.

262
00:20:52,340 --> 00:20:54,060
Andy Piper: Therefore, they will not give up.

263
00:20:54,060 --> 00:21:03,880
Andy Piper: And everybody is punished by Anubis because everybody ends up having to wait to see the webpage and to prove that they're human and get the data.

264
00:21:03,880 --> 00:21:06,080
Andy Piper: Okay, so we wanted to just end on a couple of fun things.

265
00:21:06,080 --> 00:21:07,520
Andy Piper: It's Tiny Awards season.

266
00:21:07,520 --> 00:21:08,080
Michael Martine: Hooray!

267
00:21:08,080 --> 00:21:11,120
Andy Piper: The Tiny Awards website is up.

268
00:21:11,120 --> 00:21:13,260
Andy Piper: I believe there are nine sites nominated.

269
00:21:13,260 --> 00:21:15,840
Michael Martine: They don't have a podcast category now, do they?

270
00:21:15,840 --> 00:21:19,260
Andy Piper: I don't think they do.

271
00:21:19,260 --> 00:21:23,320
Andy Piper: One, two, three, four, five, six, seven, eight, nine.

272
00:21:23,320 --> 00:21:27,280
Andy Piper: Yes, there are nine sites, sites of the year, covering a whole range of things.

273
00:21:27,280 --> 00:21:28,660
Andy Piper: They're really fun.

274
00:21:28,660 --> 00:21:32,020
Andy Piper: And a couple of them that I pulled out for our show purposes,

275
00:21:32,220 --> 00:21:37,160
Andy Piper: one of them is from our friend or the website that's shown up a number of times on our show,

276
00:21:37,160 --> 00:21:39,800
UNKNOWN: neal.fun, N-E-A-L.fun.

277
00:21:39,800 --> 00:21:44,920
Andy Piper: And he has a new site called Cursor Camp, a campsite that you can go visit with your pointer.

278
00:21:44,920 --> 00:21:46,700
Andy Piper: And it's collaborative.

279
00:21:47,700 --> 00:21:53,980
Andy Piper: Every pointer or person on the site comes with a little flag representing the country that they're from.

280
00:21:53,980 --> 00:21:58,340
Andy Piper: And you can explore this little website with your cursor.

281
00:21:58,340 --> 00:21:59,200
Andy Piper: Scroll around.

282
00:21:59,200 --> 00:22:01,120
Andy Piper: Interact with objects.

283
00:22:01,120 --> 00:22:05,520
Andy Piper: There's a camp stage with a boombox.

284
00:22:05,520 --> 00:22:09,320
Andy Piper: There's a soccer field where you can play kind of table football.

285
00:22:09,320 --> 00:22:10,120
Andy Piper: There's a treehouse.

286
00:22:10,120 --> 00:22:10,720
Andy Piper: There's a waterfall.

287
00:22:10,720 --> 00:22:11,840
Andy Piper: It's delightful.

288
00:22:11,840 --> 00:22:12,420
Andy Piper: It's lovely.

289
00:22:12,900 --> 00:22:13,660
Andy Piper: You can click on things.

290
00:22:13,660 --> 00:22:14,680
Andy Piper: You can go inside a cave.

291
00:22:14,680 --> 00:22:16,040
Andy Piper: I don't know if you found a cave, Michael.

292
00:22:16,040 --> 00:22:17,080
Andy Piper: Get your torch out.

293
00:22:17,080 --> 00:22:23,040
Andy Piper: Click into a cave and your pointer gets a little torch and you can look around.

294
00:22:23,040 --> 00:22:25,140
Andy Piper: So that's fun.

295
00:22:25,140 --> 00:22:25,880
Andy Piper: I like it.

296
00:22:25,880 --> 00:22:26,480
Michael Martine: I like it.

297
00:22:27,260 --> 00:22:30,080
Andy Piper: And then the other one, which I really liked.

298
00:22:30,080 --> 00:22:33,240
Andy Piper: I mean, I liked all nine sites, but I had to vote for one.

299
00:22:33,240 --> 00:22:35,040
Andy Piper: I'm not going to tell you which one I did vote for.

300
00:22:35,040 --> 00:22:42,900
Andy Piper: But the other one that I wanted to bring to the attention of my co-hosts and our audience is Hallucinate.

301
00:22:42,900 --> 00:22:49,860
Andy Piper: And Hallucinate is a massively multiplayer online rave, which also runs in the browser.

302
00:22:49,860 --> 00:22:51,040
Andy Piper: All of these things are in the browser.

303
00:22:51,600 --> 00:22:53,360
Andy Piper: And in this case, Hallucinate will load.

304
00:22:53,360 --> 00:22:57,260
Andy Piper: You have to type in a name for yourself to enter the field.

305
00:22:57,260 --> 00:23:06,600
Andy Piper: And you end up in a not entirely Second Life, dissimilar looking, kind of blocky, more kind of Minecraft, I suppose, looking field.

306
00:23:06,600 --> 00:23:17,100
Andy Piper: Where you can rave and hang out in a 3D space with other avatars.

307
00:23:17,100 --> 00:23:20,220
Andy Piper: There's music playing on the stage.

308
00:23:20,300 --> 00:23:22,360
Andy Piper: It's kind of like a big festival going on.

309
00:23:22,360 --> 00:23:23,700
Andy Piper: You can customize your avatar.

310
00:23:23,700 --> 00:23:26,040
Andy Piper: And that's all there is to it.

311
00:23:26,040 --> 00:23:26,320
Michael Martine: Oh, yeah.

312
00:23:26,320 --> 00:23:26,920
Andy Piper: It's super fun.

313
00:23:26,920 --> 00:23:27,840
Michael Martine: That's a good deal.

314
00:23:27,840 --> 00:23:30,060
Michael Martine: I'm just raving here just a little bit.

315
00:23:30,060 --> 00:23:31,400
Michael Martine: All right.

316
00:23:31,400 --> 00:23:38,180
Michael Martine: Well, if our friends that are listening to this want to rave with us, what should we suggest they do, do you think, Andy?

317
00:23:38,180 --> 00:23:41,240
Andy Piper: I would check us out on the Fediverse.

318
00:23:41,240 --> 00:23:50,600
Andy Piper: We have our games at work at games at work.biz or games at work at underscore biz at mastodon dot social are both handles where you can find us.

319
00:23:50,600 --> 00:23:53,760
Andy Piper: You can come to our website, games at work.biz, which is where you'll find everything.

320
00:23:53,760 --> 00:23:57,660
Andy Piper: All our show notes, all of the details of all the stuff we've been talking about.

321
00:23:57,660 --> 00:23:58,860
Andy Piper: You go and explore it yourself.

322
00:23:59,480 --> 00:24:00,840
Andy Piper: Give those sites your clicks.

323
00:24:00,840 --> 00:24:01,820
Andy Piper: Give them your attention.

324
00:24:01,820 --> 00:24:04,220
Andy Piper: And by all means, let us know what you think.

325
00:24:04,220 --> 00:24:05,780
Andy Piper: We always love to hear from our listeners.

326
00:24:05,780 --> 00:24:06,220
Michael Martine: Absolutely.

327
00:24:06,220 --> 00:24:10,940
Michael Martine: So we'll see you again here next time, next week, right back at games at work.biz.

328
00:24:10,940 --> 00:24:11,980
Michael Martine: Thanks, everybody.

329
00:24:11,980 --> 00:24:12,920
Michael Martine: See you.

330
00:24:12,920 --> 00:24:13,060
Michael Martine: Bye.

331
00:24:13,440 --> 00:24:13,740
Andy Piper: See ya.

332
00:24:13,740 --> 00:24:20,040
Intro / Outtro: You've been listening to games@work.biz, the podcast about gaming technology and play.

333
00:24:20,040 --> 00:24:26,300
Intro / Outtro: We are part of the Bluberry podcasting network, and we'd like to thank the band, Random Encounters for their song, Big Blue.

334
00:24:26,300 --> 00:24:30,540
Intro / Outtro: You can follow us at our website at gamesatwork.biz.
