October 1, 2026
2026-10-01
https://www.youtube.com/watch?v=Vc3ZLTM5a7o

[0:20] We go live in 10.
[0:34] Every morning the future moves faster.
[0:39] Every morning the future moves faster. The models, the money, the power, and
[0:43] The models, the money, the power, and
[0:43] The models, the money, the power, and the people building what comes next.
[1:18] from Austin, Texas. to the world.
[1:25] AI news, real context. This is the
[1:29] AI news, real context. This is the Austin Wilson Show. I'm Austin Wilson,
[1:32] Austin Wilson Show. I'm Austin Wilson,
[1:32] Austin Wilson Show. I'm Austin Wilson, your human in the loop.
[1:42] Welcome to the Austin Wilson Show. My
[1:44] Welcome to the Austin Wilson Show. My name is Austin Wilson. Today we're going
[1:47] name is Austin Wilson. Today we're going
[1:47] name is Austin Wilson. Today we're going to cover the most important AI news that
[1:50] to cover the most important AI news that
[1:50] to cover the most important AI news that will impact you today. Coming to you
[1:52] will impact you today. Coming to you
[1:52] will impact you today. Coming to you live from the AI
[1:55] live from the AI
[1:55] live from the AI capital, the live AI capital of the
[1:58] capital, the live AI capital of the
[1:58] capital, the live AI capital of the world, Austin, Texas. Uh we've got a
[2:01] world, Austin, Texas. Uh we've got a
[2:01] world, Austin, Texas. Uh we've got a really fun docket today. Uh really
[2:03] really fun docket today. Uh really
[2:03] really fun docket today. Uh really interesting stories. Hopefully they have
[2:04] interesting stories. Hopefully they have
[2:04] interesting stories. Hopefully they have an impact on you. You learn something
[2:06] an impact on you. You learn something
[2:06] an impact on you. You learn something new. We have some fun. Uh let's get into
[2:09] new. We have some fun. Uh let's get into
[2:09] new. We have some fun. Uh let's get into it. Uh we're first going to cover a
[2:12] it. Uh we're first going to cover a
[2:12] it. Uh we're first going to cover a breaking story out of uh Financial
[2:16] breaking story out of uh Financial
[2:16] breaking story out of uh Financial Times. They are reporting that there's
[2:18] Times. They are reporting that there's
[2:18] Times. They are reporting that there's been a another Open AI agent incident
[2:22] been a another Open AI agent incident
[2:22] been a another Open AI agent incident where uh OpenAI agents uh went a little
[2:26] where uh OpenAI agents uh went a little
[2:26] where uh OpenAI agents uh went a little bit rogue. Um, looks like they uh broke
[2:30] bit rogue. Um, looks like they uh broke
[2:30] bit rogue. Um, looks like they uh broke out and accessed websites including the
[2:34] out and accessed websites including the
[2:34] out and accessed websites including the CDC, the SEC, the Mayo Clinic uh and
[2:37] CDC, the SEC, the Mayo Clinic uh and
[2:37] CDC, the SEC, the Mayo Clinic uh and tried to cover their tracks in doing so.
[2:41] tried to cover their tracks in doing so.
[2:41] tried to cover their tracks in doing so. According to digital forensic firm
[2:44] According to digital forensic firm
[2:44] According to digital forensic firm asymmetric security, the agents use new
[2:48] asymmetric security, the agents use new
[2:48] asymmetric security, the agents use new methods to erase records or make them
[2:51] methods to erase records or make them
[2:51] methods to erase records or make them inaccessible. Uh they set up temporary
[2:53] inaccessible. Uh they set up temporary
[2:53] inaccessible. Uh they set up temporary email inboxes and private accounts to
[2:57] email inboxes and private accounts to
[2:57] email inboxes and private accounts to download the data.
[3:00] download the data.
[3:00] download the data. Wow. Um going rogue on us.
[3:04] Wow. Um going rogue on us.
[3:04] Wow. Um going rogue on us. Not the first time we've seen this. Um
[3:07] Not the first time we've seen this. Um
[3:07] Not the first time we've seen this. Um it it it almost seems like stale news
[3:10] it it it almost seems like stale news
[3:10] it it it almost seems like stale news now. That's just a normal thing. Agents
[3:12] now. That's just a normal thing. Agents
[3:12] now. That's just a normal thing. Agents break out. I don't think that's too
[3:14] break out. I don't think that's too
[3:14] break out. I don't think that's too normal and it it's a little bit
[3:16] normal and it it's a little bit
[3:16] normal and it it's a little bit shocking. That being said, um we don't
[3:20] shocking. That being said, um we don't
[3:20] shocking. That being said, um we don't know all the details around the story
[3:22] know all the details around the story
[3:22] know all the details around the story and it'll be interesting to see as they
[3:25] and it'll be interesting to see as they
[3:25] and it'll be interesting to see as they come out uh what what were the
[3:27] come out uh what what were the
[3:27] come out uh what what were the circumstances around this situation.
[3:30] circumstances around this situation.
[3:30] circumstances around this situation. It's kind of interesting to the the
[3:32] It's kind of interesting to the the
[3:32] It's kind of interesting to the the reporting
[3:34] reporting
[3:34] reporting the way that I read it and we can get
[3:36] the way that I read it and we can get
[3:36] the way that I read it and we can get into this. They uh they access 55
[3:39] into this. They uh they access 55
[3:39] into this. They uh they access 55 websites but the reporters highlight the
[3:42] websites but the reporters highlight the
[3:42] websites but the reporters highlight the kind of scary one or or the curious
[3:44] kind of scary one or or the curious
[3:44] kind of scary one or or the curious ones. they access the CDC and Mayo
[3:46] ones. they access the CDC and Mayo
[3:46] ones. they access the CDC and Mayo Clinic and SEC, but why? Um, so yeah, we
[3:50] Clinic and SEC, but why? Um, so yeah, we
[3:50] Clinic and SEC, but why? Um, so yeah, we who knows where we're at on this. Uh,
[3:52] who knows where we're at on this. Uh,
[3:52] who knows where we're at on this. Uh, it's still early, but it's interesting
[3:54] it's still early, but it's interesting
[3:54] it's still early, but it's interesting to see what's going on. Uh, OpenAI says
[3:57] to see what's going on. Uh, OpenAI says
[3:57] to see what's going on. Uh, OpenAI says it's reviewing misaligned model activity
[4:00] it's reviewing misaligned model activity
[4:00] it's reviewing misaligned model activity and that most of that was detected um
[4:04] and that most of that was detected um
[4:04] and that most of that was detected um was routine research on public content.
[4:07] was routine research on public content.
[4:07] was routine research on public content. Okay, so maybe that's a little bit re
[4:10] Okay, so maybe that's a little bit re
[4:10] Okay, so maybe that's a little bit re reassuring. Um, it's also reassuring to
[4:13] reassuring. Um, it's also reassuring to
[4:13] reassuring. Um, it's also reassuring to me that they have an outside firm
[4:15] me that they have an outside firm
[4:15] me that they have an outside firm auditing this. We don't have a date yet
[4:17] auditing this. We don't have a date yet
[4:17] auditing this. We don't have a date yet either, at least I did not see one on
[4:19] either, at least I did not see one on
[4:19] either, at least I did not see one on when this incident occurred. Um, we
[4:22] when this incident occurred. Um, we
[4:22] when this incident occurred. Um, we don't know quite how old uh this is.
[4:26] don't know quite how old uh this is.
[4:26] don't know quite how old uh this is. you know, we're going to, I think,
[4:29] you know, we're going to, I think,
[4:29] you know, we're going to, I think, continue to see, especially as this is
[4:33] continue to see, especially as this is
[4:33] continue to see, especially as this is kind of a hot topic is agents breaking
[4:36] kind of a hot topic is agents breaking
[4:36] kind of a hot topic is agents breaking out of the sandbox and going and doing
[4:39] out of the sandbox and going and doing
[4:39] out of the sandbox and going and doing their own things. Um it it it I I can
[4:43] their own things. Um it it it I I can
[4:43] their own things. Um it it it I I can almost guarantee you that we're going to
[4:45] almost guarantee you that we're going to
[4:45] almost guarantee you that we're going to continue to see these stories, but it's
[4:47] continue to see these stories, but it's
[4:47] continue to see these stories, but it's going to become more and more of a
[4:48] going to become more and more of a
[4:48] going to become more and more of a nothing burger as um
[4:51] nothing burger as um
[4:51] nothing burger as um the guard rails, the protections, the
[4:54] the guard rails, the protections, the
[4:54] the guard rails, the protections, the the safety around
[4:56] the safety around
[4:56] the safety around uh these agent runs and the the research
[4:59] uh these agent runs and the the research
[4:59] uh these agent runs and the the research that these uh Frontier AI companies are
[5:02] that these uh Frontier AI companies are
[5:02] that these uh Frontier AI companies are doing uh becomes better and better, more
[5:05] doing uh becomes better and better, more
[5:05] doing uh becomes better and better, more robust. um we might see these breakouts,
[5:07] robust. um we might see these breakouts,
[5:08] robust. um we might see these breakouts, but uh we're going to have um I think
[5:12] but uh we're going to have um I think
[5:12] but uh we're going to have um I think strong controls around that, strong
[5:14] strong controls around that, strong
[5:14] strong controls around that, strong monitoring and being able to understand
[5:16] monitoring and being able to understand
[5:16] monitoring and being able to understand what's going on. And and the goal here
[5:18] what's going on. And and the goal here
[5:18] what's going on. And and the goal here hopefully too is and I know what they're
[5:20] hopefully too is and I know what they're
[5:20] hopefully too is and I know what they're doing. They're testing the agents to see
[5:22] doing. They're testing the agents to see
[5:22] doing. They're testing the agents to see the capabilities, how far can they push
[5:24] the capabilities, how far can they push
[5:24] the capabilities, how far can they push it, what happens when they open things
[5:27] it, what happens when they open things
[5:27] it, what happens when they open things up as much as possible, but they need to
[5:29] up as much as possible, but they need to
[5:29] up as much as possible, but they need to do a better job of paying attention to
[5:31] do a better job of paying attention to
[5:31] do a better job of paying attention to what's going on if it actually is
[5:33] what's going on if it actually is
[5:33] what's going on if it actually is dangerous. Now, we don't know that this
[5:34] dangerous. Now, we don't know that this
[5:34] dangerous. Now, we don't know that this is this is dangerous yet. Um, it's still
[5:37] is this is dangerous yet. Um, it's still
[5:37] is this is dangerous yet. Um, it's still early, but um I I can almost guarantee
[5:40] early, but um I I can almost guarantee
[5:40] early, but um I I can almost guarantee you we'll continue to see stories like
[5:42] you we'll continue to see stories like
[5:42] you we'll continue to see stories like this. AI agent goes rogue, breaks out of
[5:45] this. AI agent goes rogue, breaks out of
[5:45] this. AI agent goes rogue, breaks out of sandbox.
[5:47] sandbox.
[5:47] sandbox. To to me at this point, it's nothing to
[5:49] To to me at this point, it's nothing to
[5:49] To to me at this point, it's nothing to be um concerned about. Um, you know,
[5:55] be um concerned about. Um, you know,
[5:55] be um concerned about. Um, you know, maybe the more concerning part is that
[5:57] maybe the more concerning part is that
[5:57] maybe the more concerning part is that these are closed labs. So, we don't have
[6:00] these are closed labs. So, we don't have
[6:00] these are closed labs. So, we don't have the uh trace logs on what the agents
[6:03] the uh trace logs on what the agents
[6:03] the uh trace logs on what the agents actually did. We don't know the chain of
[6:05] actually did. We don't know the chain of
[6:05] actually did. We don't know the chain of thought. We don't know what data they
[6:08] thought. We don't know what data they
[6:08] thought. We don't know what data they actually accessed. Um, and that's where
[6:12] actually accessed. Um, and that's where
[6:12] actually accessed. Um, and that's where maybe it's a little bit concerning and
[6:14] maybe it's a little bit concerning and
[6:14] maybe it's a little bit concerning and where the the open- source is um, uh,
[6:17] where the the open- source is um, uh,
[6:17] where the the open- source is um, uh, there's an argument to may be made for
[6:20] there's an argument to may be made for
[6:20] there's an argument to may be made for open software where maybe we have
[6:22] open software where maybe we have
[6:22] open software where maybe we have visibility into what's going on and what
[6:24] visibility into what's going on and what
[6:24] visibility into what's going on and what what data is being accessed.
[6:27] what data is being accessed.
[6:27] what data is being accessed. Moving on, uh, exciting news coming out
[6:30] Moving on, uh, exciting news coming out
[6:30] Moving on, uh, exciting news coming out of Google. They unveiled their Gemini 4
[6:33] of Google. They unveiled their Gemini 4
[6:33] of Google. They unveiled their Gemini 4 Argon
[6:35] Argon
[6:35] Argon model uh calling it their most capable
[6:38] model uh calling it their most capable
[6:38] model uh calling it their most capable model yet. For now, uh it only goes out
[6:41] model yet. For now, uh it only goes out
[6:42] model yet. For now, uh it only goes out to a vetted group of cyber security
[6:44] to a vetted group of cyber security
[6:44] to a vetted group of cyber security defenders and Google says it's taking
[6:46] defenders and Google says it's taking
[6:46] defenders and Google says it's taking part in the government's voluntary
[6:49] part in the government's voluntary
[6:49] part in the government's voluntary pre-release review. uh introductory
[6:52] pre-release review. uh introductory
[6:52] pre-release review. uh introductory pricing, which is this is very
[6:55] pricing, which is this is very
[6:55] pricing, which is this is very competitive pricing. $2 per million
[6:57] competitive pricing. $2 per million
[6:57] competitive pricing. $2 per million input tokens and $10 per million output
[7:00] input tokens and $10 per million output
[7:00] input tokens and $10 per million output tokens. Um very competitive, very
[7:03] tokens. Um very competitive, very
[7:03] tokens. Um very competitive, very efficient. Great to see this coming out
[7:05] efficient. Great to see this coming out
[7:05] efficient. Great to see this coming out of Google. Uh they have been trailing
[7:09] of Google. Uh they have been trailing
[7:09] of Google. Uh they have been trailing behind everybody for quite some time.
[7:12] behind everybody for quite some time.
[7:12] behind everybody for quite some time. Um, Google had a big splash in summer of
[7:15] Um, Google had a big splash in summer of
[7:15] Um, Google had a big splash in summer of 2025 with Nano Banana and VO3, their
[7:21] 2025 with Nano Banana and VO3, their
[7:21] 2025 with Nano Banana and VO3, their video model, doing some really cool
[7:23] video model, doing some really cool
[7:23] video model, doing some really cool things on uh, kind of the creative side,
[7:27] things on uh, kind of the creative side,
[7:27] things on uh, kind of the creative side, but we didn't see as much on the
[7:29] but we didn't see as much on the
[7:29] but we didn't see as much on the intelligent side. They've been lagging
[7:31] intelligent side. They've been lagging
[7:31] intelligent side. They've been lagging behind OpenAI and Anthropic and uh
[7:35] behind OpenAI and Anthropic and uh
[7:35] behind OpenAI and Anthropic and uh SpaceX AI and even Metamuse, Ozuck
[7:40] SpaceX AI and even Metamuse, Ozuck
[7:40] SpaceX AI and even Metamuse, Ozuck coming out with uh Muse this summer and
[7:44] coming out with uh Muse this summer and
[7:44] coming out with uh Muse this summer and and seeing the capabilities there. Um
[7:47] and seeing the capabilities there. Um
[7:47] and seeing the capabilities there. Um Google was supposed to launch their
[7:48] Google was supposed to launch their
[7:48] Google was supposed to launch their Gemini 3.5 Pro. Um it looks like they
[7:52] Gemini 3.5 Pro. Um it looks like they
[7:52] Gemini 3.5 Pro. Um it looks like they tabled that to focus on Argon um which
[7:57] tabled that to focus on Argon um which
[7:57] tabled that to focus on Argon um which apparently is a really good thing. Um,
[7:59] apparently is a really good thing. Um,
[7:59] apparently is a really good thing. Um, Sundar Pachai, the the CEO of Google,
[8:03] Sundar Pachai, the the CEO of Google,
[8:03] Sundar Pachai, the the CEO of Google, uh, retweeted some preliminary
[8:05] uh, retweeted some preliminary
[8:05] uh, retweeted some preliminary benchmarks. If you're not familiar, uh,
[8:08] benchmarks. If you're not familiar, uh,
[8:08] benchmarks. If you're not familiar, uh, the the Frontier model companies use
[8:11] the the Frontier model companies use
[8:11] the the Frontier model companies use these benchmarks to compare models
[8:14] these benchmarks to compare models
[8:14] these benchmarks to compare models relatively to one another.
[8:17] relatively to one another.
[8:17] relatively to one another. It's hard to tell really what the
[8:19] It's hard to tell really what the
[8:19] It's hard to tell really what the capabilities are on these benchmarks and
[8:22] capabilities are on these benchmarks and
[8:22] capabilities are on these benchmarks and how they measure them. Um but in this
[8:24] how they measure them. Um but in this
[8:24] how they measure them. Um but in this case uh Argon um almost
[8:29] case uh Argon um almost
[8:29] case uh Argon um almost beat everybody uh across the board uh on
[8:33] beat everybody uh across the board uh on
[8:33] beat everybody uh across the board uh on the the benchmarks that they did
[8:35] the the benchmarks that they did
[8:35] the the benchmarks that they did release. And so it looks really good.
[8:37] release. And so it looks really good.
[8:37] release. And so it looks really good. This is this is really good coming out
[8:39] This is this is really good coming out
[8:39] This is this is really good coming out of Google. Uh congratulations
[8:42] of Google. Uh congratulations
[8:42] of Google. Uh congratulations to the Google team and um happy to see
[8:45] to the Google team and um happy to see
[8:45] to the Google team and um happy to see that we have another big player in the
[8:48] that we have another big player in the
[8:48] that we have another big player in the race. Interestingly enough, I was
[8:51] race. Interestingly enough, I was
[8:51] race. Interestingly enough, I was talking to somebody yesterday uh out of
[8:53] talking to somebody yesterday uh out of
[8:53] talking to somebody yesterday uh out of Norway that said that Google is 5% of
[8:58] Norway that said that Google is 5% of
[8:58] Norway that said that Google is 5% of Norway's electricity consumption running
[9:02] Norway's electricity consumption running
[9:02] Norway's electricity consumption running their data centers, which is uh amazing.
[9:04] their data centers, which is uh amazing.
[9:04] their data centers, which is uh amazing. And then um I saw some somewhere else,
[9:07] And then um I saw some somewhere else,
[9:07] And then um I saw some somewhere else, apparently Google is maybe a third of
[9:09] apparently Google is maybe a third of
[9:09] apparently Google is maybe a third of the world's compute or up to or over
[9:12] the world's compute or up to or over
[9:12] the world's compute or up to or over that. Crazy. Google's massive. Great to
[9:15] that. Crazy. Google's massive. Great to
[9:15] that. Crazy. Google's massive. Great to see that they're in um uh in the fight
[9:19] see that they're in um uh in the fight
[9:19] see that they're in um uh in the fight again and hopefully this brings more
[9:22] again and hopefully this brings more
[9:22] again and hopefully this brings more intelligence, more capabilities to more
[9:25] intelligence, more capabilities to more
[9:25] intelligence, more capabilities to more people. Um I I think what's exciting
[9:29] people. Um I I think what's exciting
[9:29] people. Um I I think what's exciting about this number one is the
[9:30] about this number one is the
[9:30] about this number one is the competition. Like the more competition
[9:33] competition. Like the more competition
[9:33] competition. Like the more competition we have in this space, I think the
[9:35] we have in this space, I think the
[9:35] we have in this space, I think the better um the more intelligence um and
[9:39] better um the more intelligence um and
[9:39] better um the more intelligence um and and capability is going to spread to
[9:41] and capability is going to spread to
[9:41] and capability is going to spread to more people. Um, and then just the the
[9:44] more people. Um, and then just the the
[9:44] more people. Um, and then just the the advancement and and not just
[9:46] advancement and and not just
[9:46] advancement and and not just productivity, but what everybody's able
[9:48] productivity, but what everybody's able
[9:48] productivity, but what everybody's able to do. And hopefully um uh we see more
[9:51] to do. And hopefully um uh we see more
[9:51] to do. And hopefully um uh we see more and more lives improved because of this.
[9:54] and more lives improved because of this.
[9:54] and more lives improved because of this. Um, and and so that's that's why it's
[9:56] Um, and and so that's that's why it's
[9:56] Um, and and so that's that's why it's super exciting. That's why the
[9:57] super exciting. That's why the
[9:57] super exciting. That's why the competition is exciting. From a um
[10:00] competition is exciting. From a um
[10:00] competition is exciting. From a um economic standpoint, I think Google is
[10:02] economic standpoint, I think Google is
[10:02] economic standpoint, I think Google is positioned extremely well. Their
[10:05] positioned extremely well. Their
[10:05] positioned extremely well. Their distribution is insane and they've
[10:07] distribution is insane and they've
[10:07] distribution is insane and they've historically done a really really good
[10:09] historically done a really really good
[10:09] historically done a really really good job at product development. And you
[10:12] job at product development. And you
[10:12] job at product development. And you know, you look across the board and and
[10:13] know, you look across the board and and
[10:14] know, you look across the board and and their surface area, you know, I think
[10:16] their surface area, you know, I think
[10:16] their surface area, you know, I think the average per person probably touches
[10:18] the average per person probably touches
[10:18] the average per person probably touches a Google product uh a dozen times a day
[10:21] a Google product uh a dozen times a day
[10:21] a Google product uh a dozen times a day at least, if not more, which is amazing.
[10:24] at least, if not more, which is amazing.
[10:24] at least, if not more, which is amazing. And um I, you know, I think and my
[10:27] And um I, you know, I think and my
[10:27] And um I, you know, I think and my perspective has been that their um kind
[10:31] perspective has been that their um kind
[10:31] perspective has been that their um kind of AI strategy has been twofold. And I
[10:34] of AI strategy has been twofold. And I
[10:34] of AI strategy has been twofold. And I think it's actually really interesting,
[10:36] think it's actually really interesting,
[10:36] think it's actually really interesting, especially after um the the change in
[10:39] especially after um the the change in
[10:39] especially after um the the change in leadership at Deep Mind and seeing Demis
[10:42] leadership at Deep Mind and seeing Demis
[10:42] leadership at Deep Mind and seeing Demis kind of uh move up or move out, I don't
[10:46] kind of uh move up or move out, I don't
[10:46] kind of uh move up or move out, I don't know, and and seeing some talent changes
[10:48] know, and and seeing some talent changes
[10:48] know, and and seeing some talent changes over there. Uh it it's interesting to
[10:51] over there. Uh it it's interesting to
[10:51] over there. Uh it it's interesting to see this release and and really happy
[10:53] see this release and and really happy
[10:53] see this release and and really happy for them. But I think Google's strategy
[10:55] for them. But I think Google's strategy
[10:55] for them. But I think Google's strategy has been twofold. one um the data center
[10:59] has been twofold. one um the data center
[10:59] has been twofold. one um the data center game, you know, Google Cloud is uh a
[11:02] game, you know, Google Cloud is uh a
[11:02] game, you know, Google Cloud is uh a powerhouse
[11:04] powerhouse
[11:04] powerhouse um in the the cloud compute world. And
[11:07] um in the the cloud compute world. And
[11:07] um in the the cloud compute world. And as we see, um you know, more and more
[11:11] as we see, um you know, more and more
[11:11] as we see, um you know, more and more entities quoting or pitching these half
[11:14] entities quoting or pitching these half
[11:14] entities quoting or pitching these half a trillion dollar compute packages, man,
[11:17] a trillion dollar compute packages, man,
[11:18] a trillion dollar compute packages, man, what a great game to be in. I'd love to
[11:20] what a great game to be in. I'd love to
[11:20] what a great game to be in. I'd love to be Google, right? Um so I think they've
[11:22] be Google, right? Um so I think they've
[11:22] be Google, right? Um so I think they've been playing that game, saying, "Hey,
[11:23] been playing that game, saying, "Hey,
[11:23] been playing that game, saying, "Hey, you know what? We'll run your models for
[11:26] you know what? We'll run your models for
[11:26] you know what? We'll run your models for you. We don't necessarily need to be the
[11:28] you. We don't necessarily need to be the
[11:28] you. We don't necessarily need to be the most intelligent. Well, wow. Look at
[11:30] most intelligent. Well, wow. Look at
[11:30] most intelligent. Well, wow. Look at this. Kind of out of left field. Um,
[11:33] this. Kind of out of left field. Um,
[11:33] this. Kind of out of left field. Um, they are right there in the intelligence
[11:34] they are right there in the intelligence
[11:34] they are right there in the intelligence game. So, uh, excited to see that.
[11:36] game. So, uh, excited to see that.
[11:36] game. So, uh, excited to see that. Excited to see where this goes. I think
[11:39] Excited to see where this goes. I think
[11:39] Excited to see where this goes. I think and and my hunch on this is that they go
[11:42] and and my hunch on this is that they go
[11:42] and and my hunch on this is that they go for the efficiency play, the economical
[11:45] for the efficiency play, the economical
[11:45] for the efficiency play, the economical play. um they're not you know they'll
[11:49] play. um they're not you know they'll
[11:49] play. um they're not you know they'll probably make available but they're
[11:50] probably make available but they're
[11:50] probably make available but they're they're probably not targeting um you
[11:54] they're probably not targeting um you
[11:54] they're probably not targeting um you know the the super high spenders they
[11:56] know the the super high spenders they
[11:56] know the the super high spenders they they're targeting the consumers um and
[12:00] they're targeting the consumers um and
[12:00] they're targeting the consumers um and enabling consumers to use AI across the
[12:03] enabling consumers to use AI across the
[12:03] enabling consumers to use AI across the board across all their you know Google
[12:07] board across all their you know Google
[12:07] board across all their you know Google products. when we look at um mobile or
[12:11] products. when we look at um mobile or
[12:11] products. when we look at um mobile or uh Gmail, the Google Workspace Suite, um
[12:15] uh Gmail, the Google Workspace Suite, um
[12:15] uh Gmail, the Google Workspace Suite, um we look across obviously search,
[12:17] we look across obviously search,
[12:17] we look across obviously search, YouTube, um we're going to see more and
[12:20] YouTube, um we're going to see more and
[12:20] YouTube, um we're going to see more and more of these elements come into play
[12:21] more of these elements come into play
[12:21] more of these elements come into play here, which is interesting. Again,
[12:23] here, which is interesting. Again,
[12:23] here, which is interesting. Again, congrats to Google on the release.
[12:25] congrats to Google on the release.
[12:26] congrats to Google on the release. Excited to test it out.
[12:28] Excited to test it out.
[12:28] Excited to test it out. Uh, in government news, moving over to
[12:32] Uh, in government news, moving over to
[12:32] Uh, in government news, moving over to California, the West Coast, we've got
[12:35] California, the West Coast, we've got
[12:35] California, the West Coast, we've got Gavin Newsome. Um, yesterday he signed
[12:40] Gavin Newsome. Um, yesterday he signed
[12:40] Gavin Newsome. Um, yesterday he signed uh new AI laws. These are all centered
[12:43] uh new AI laws. These are all centered
[12:43] uh new AI laws. These are all centered around um employers and what employers
[12:46] around um employers and what employers
[12:46] around um employers and what employers can do in regards to AI. Um, employers
[12:51] can do in regards to AI. Um, employers
[12:51] can do in regards to AI. Um, employers can't use AI alone to fire workers.
[12:54] can't use AI alone to fire workers.
[12:54] can't use AI alone to fire workers. can't read workers emotions through
[12:56] can't read workers emotions through
[12:56] can't read workers emotions through biometrics data and they must give
[12:59] biometrics data and they must give
[13:00] biometrics data and they must give written notice when AI drives mass
[13:03] written notice when AI drives mass
[13:03] written notice when AI drives mass layoffs. Um he also ordered state
[13:05] layoffs. Um he also ordered state
[13:05] layoffs. Um he also ordered state agencies to keep saying artificial
[13:08] agencies to keep saying artificial
[13:08] agencies to keep saying artificial intelligence. Uh the OG AI baby um as
[13:13] intelligence. Uh the OG AI baby um as
[13:13] intelligence. Uh the OG AI baby um as opposed to Trump's news super
[13:15] opposed to Trump's news super
[13:15] opposed to Trump's news super intelligence. Wonderful. Love to see
[13:18] intelligence. Wonderful. Love to see
[13:18] intelligence. Wonderful. Love to see that. Um, you know, Washington 2 days
[13:22] that. Um, you know, Washington 2 days
[13:22] that. Um, you know, Washington 2 days ago had kind of their voluntary pledge.
[13:25] ago had kind of their voluntary pledge.
[13:25] ago had kind of their voluntary pledge. Um, now California, the largest state
[13:28] Um, now California, the largest state
[13:28] Um, now California, the largest state economy, uh, has created some of their
[13:30] economy, uh, has created some of their
[13:30] economy, uh, has created some of their own binding rules. Really interesting to
[13:33] own binding rules. Really interesting to
[13:33] own binding rules. Really interesting to to contrast these. Um, Newsome said AI
[13:37] to contrast these. Um, Newsome said AI
[13:37] to contrast these. Um, Newsome said AI should expand opportunity, not come at
[13:39] should expand opportunity, not come at
[13:39] should expand opportunity, not come at the expense of workers and families. Um,
[13:42] the expense of workers and families. Um,
[13:42] the expense of workers and families. Um, the the governor's office lists 13
[13:45] the the governor's office lists 13
[13:45] the the governor's office lists 13 signed bills. Uh and in total, this is
[13:50] signed bills. Uh and in total, this is
[13:50] signed bills. Uh and in total, this is uh Gavin Newsome's last from from the
[13:52] uh Gavin Newsome's last from from the
[13:52] uh Gavin Newsome's last from from the way I understand it, his last uh
[13:55] way I understand it, his last uh
[13:55] way I understand it, his last uh official
[13:57] official
[13:57] official um not maybe not his last official act,
[13:59] um not maybe not his last official act,
[13:59] um not maybe not his last official act, but the yesterday was his last day to be
[14:01] but the yesterday was his last day to be
[14:01] but the yesterday was his last day to be able to sign acts into law, and this was
[14:04] able to sign acts into law, and this was
[14:04] able to sign acts into law, and this was kind of his last move uh before he
[14:06] kind of his last move uh before he
[14:06] kind of his last move uh before he leaves office in January. Um, you know,
[14:11] leaves office in January. Um, you know,
[14:11] leaves office in January. Um, you know, my take on this is
[14:17] good for them. I I love the spirit of
[14:19] good for them. I I love the spirit of it. I love um I I love the concern
[14:23] it. I love um I I love the concern
[14:23] it. I love um I I love the concern around
[14:24] around
[14:24] around um maybe kind of protecting against
[14:27] um maybe kind of protecting against
[14:27] um maybe kind of protecting against disruption. Uh
[14:31] disruption. Uh
[14:31] disruption. Uh it's kind of hilar hilarious, but I'm
[14:33] it's kind of hilar hilarious, but I'm
[14:33] it's kind of hilar hilarious, but I'm like thinking about employers using AI
[14:37] like thinking about employers using AI
[14:37] like thinking about employers using AI to fire people. That's outrageous. If
[14:41] to fire people. That's outrageous. If
[14:41] to fire people. That's outrageous. If any employer ever did that, you know, I
[14:44] any employer ever did that, you know, I
[14:44] any employer ever did that, you know, I would hope they would get lambasted in
[14:46] would hope they would get lambasted in
[14:46] would hope they would get lambasted in the media that their customers would
[14:47] the media that their customers would
[14:47] the media that their customers would leave. I can't imagine that. I I haven't
[14:51] leave. I can't imagine that. I I haven't
[14:51] leave. I can't imagine that. I I haven't heard that as a use case yet. I haven't,
[14:54] heard that as a use case yet. I haven't,
[14:54] heard that as a use case yet. I haven't, you know, haven't even really heard a
[14:56] you know, haven't even really heard a
[14:56] you know, haven't even really heard a lot of HR
[14:59] lot of HR
[14:59] lot of HR uh use cases beyond maybe automated
[15:01] uh use cases beyond maybe automated
[15:01] uh use cases beyond maybe automated onboarding kind of stuff. Um, but
[15:06] onboarding kind of stuff. Um, but
[15:06] onboarding kind of stuff. Um, but glad that they're they're covering
[15:08] glad that they're they're covering
[15:08] glad that they're they're covering covering that, looking out for workers.
[15:11] covering that, looking out for workers.
[15:11] covering that, looking out for workers. Um, the biometric data thing is kind of
[15:13] Um, the biometric data thing is kind of
[15:13] Um, the biometric data thing is kind of interesting as well, too. Um,
[15:15] interesting as well, too. Um,
[15:15] interesting as well, too. Um, interesting to see that. And then just
[15:17] interesting to see that. And then just
[15:17] interesting to see that. And then just hilarious. I I just love the um the
[15:19] hilarious. I I just love the um the
[15:19] hilarious. I I just love the um the petty back and forth or like uh you
[15:21] petty back and forth or like uh you
[15:21] petty back and forth or like uh you know, Trump changes it to super
[15:24] know, Trump changes it to super
[15:24] know, Trump changes it to super intelligence and Gavin's like, "No, I
[15:27] intelligence and Gavin's like, "No, I
[15:27] intelligence and Gavin's like, "No, I don't know. We're using artificial
[15:28] don't know. We're using artificial
[15:28] don't know. We're using artificial intelligence." Um you know key key point
[15:31] intelligence." Um you know key key point
[15:31] intelligence." Um you know key key point on this is um I think if we zoom out
[15:34] on this is um I think if we zoom out
[15:34] on this is um I think if we zoom out overall good news you know that the
[15:36] overall good news you know that the
[15:36] overall good news you know that the government is paying attention to this
[15:38] government is paying attention to this
[15:38] government is paying attention to this paying attention to I mean how could you
[15:41] paying attention to I mean how could you
[15:41] paying attention to I mean how could you not right but like um paying attention
[15:44] not right but like um paying attention
[15:44] not right but like um paying attention to how this impacts the economy um to me
[15:48] to how this impacts the economy um to me
[15:48] to how this impacts the economy um to me from the reading of this it doesn't
[15:49] from the reading of this it doesn't
[15:49] from the reading of this it doesn't sound like it's anything destructive or
[15:52] sound like it's anything destructive or
[15:52] sound like it's anything destructive or super restrictive from an AI standpoint
[15:55] super restrictive from an AI standpoint
[15:55] super restrictive from an AI standpoint um which is great but I would rather see
[15:59] um which is great but I would rather see
[15:59] um which is great but I would rather see the positive side of it. I you know
[16:01] the positive side of it. I you know
[16:01] the positive side of it. I you know we're we're um you know you you you go
[16:04] we're we're um you know you you you go
[16:04] we're we're um you know you you you go back to uh Charlie Mer you know Warren
[16:09] back to uh Charlie Mer you know Warren
[16:09] back to uh Charlie Mer you know Warren Buffett and you know measuring
[16:10] Buffett and you know measuring
[16:10] Buffett and you know measuring incentives and let's let's look at the
[16:12] incentives and let's let's look at the
[16:12] incentives and let's let's look at the incentives right um I would rather
[16:14] incentives right um I would rather
[16:14] incentives right um I would rather incentivize businesses to um encourage
[16:19] incentivize businesses to um encourage
[16:19] incentivize businesses to um encourage you know encourage them to put money
[16:21] you know encourage them to put money
[16:21] you know encourage them to put money into AI education and um uh improving
[16:26] into AI education and um uh improving
[16:26] into AI education and um uh improving skill sets around AI and how do we get
[16:28] skill sets around AI and how do we get
[16:28] skill sets around AI and how do we get more workers and more users um skilled
[16:32] more workers and more users um skilled
[16:32] more workers and more users um skilled up on these amazing tools and to be able
[16:34] up on these amazing tools and to be able
[16:34] up on these amazing tools and to be able to um use these proficiently and become
[16:38] to um use these proficiently and become
[16:38] to um use these proficiently and become more and more productive. I would rather
[16:40] more and more productive. I would rather
[16:40] more and more productive. I would rather see moves that way instead of um you
[16:43] see moves that way instead of um you
[16:43] see moves that way instead of um you know all these protections and and maybe
[16:45] know all these protections and and maybe
[16:45] know all these protections and and maybe some of it's based in AI doom and you
[16:49] some of it's based in AI doom and you
[16:49] some of it's based in AI doom and you know whatever it it's just it definitely
[16:51] know whatever it it's just it definitely
[16:51] know whatever it it's just it definitely is an interesting story. We'll see the
[16:53] is an interesting story. We'll see the
[16:53] is an interesting story. We'll see the impact of this uh as as the law goes
[16:56] impact of this uh as as the law goes
[16:56] impact of this uh as as the law goes into effect and um see if there's any
[17:00] into effect and um see if there's any
[17:00] into effect and um see if there's any real impact there. I'm not sure that
[17:02] real impact there. I'm not sure that
[17:02] real impact there. I'm not sure that this will make a big splash, but um uh
[17:06] this will make a big splash, but um uh
[17:06] this will make a big splash, but um uh great to see that Gavin's, you know,
[17:10] great to see that Gavin's, you know,
[17:10] great to see that Gavin's, you know, signing stuff on the way out. Love it.
[17:12] signing stuff on the way out. Love it.
[17:12] signing stuff on the way out. Love it. Uh moving on to Deep Seek. Um, if you're
[17:17] Uh moving on to Deep Seek. Um, if you're
[17:17] Uh moving on to Deep Seek. Um, if you're not familiar, don't recall, um, about a
[17:21] not familiar, don't recall, um, about a
[17:21] not familiar, don't recall, um, about a year and a half ago, almost two years
[17:22] year and a half ago, almost two years
[17:22] year and a half ago, almost two years ago, Deepseek,
[17:24] ago, Deepseek,
[17:24] ago, Deepseek, uh, and, uh, a Chinese um, maybe
[17:28] uh, and, uh, a Chinese um, maybe
[17:28] uh, and, uh, a Chinese um, maybe Frontier model company, they released a
[17:31] Frontier model company, they released a
[17:31] Frontier model company, they released a Frontier AI model that was extremely
[17:35] Frontier AI model that was extremely
[17:35] Frontier AI model that was extremely competitive with, um, you know, OpenAI
[17:38] competitive with, um, you know, OpenAI
[17:38] competitive with, um, you know, OpenAI and Anthropic at the time, and they
[17:40] and Anthropic at the time, and they
[17:40] and Anthropic at the time, and they continue to put out really, really good
[17:44] continue to put out really, really good
[17:44] continue to put out really, really good um, opensource course models. Now, that
[17:47] um, opensource course models. Now, that
[17:47] um, opensource course models. Now, that being said, they they're starting to
[17:49] being said, they they're starting to
[17:49] being said, they they're starting to charge more for the models, but that's a
[17:52] charge more for the models, but that's a
[17:52] charge more for the models, but that's a different topic. Um, what what we see
[17:54] different topic. Um, what what we see
[17:54] different topic. Um, what what we see coming out of China today is the
[17:55] coming out of China today is the
[17:55] coming out of China today is the Deepseek released six open-source
[17:58] Deepseek released six open-source
[17:58] Deepseek released six open-source software tools built for Huawei's Ascend
[18:02] software tools built for Huawei's Ascend
[18:02] software tools built for Huawei's Ascend AI chips. They include uh Tile Lang,
[18:05] AI chips. They include uh Tile Lang,
[18:05] AI chips. They include uh Tile Lang, which is a programming language DeepS
[18:08] which is a programming language DeepS
[18:08] which is a programming language DeepS pitches as an alternative to Nvidia's
[18:11] pitches as an alternative to Nvidia's
[18:11] pitches as an alternative to Nvidia's CUDA. Um, Huawei backed the effort and
[18:15] CUDA. Um, Huawei backed the effort and
[18:15] CUDA. Um, Huawei backed the effort and the two tuned a super mode of
[18:19] the two tuned a super mode of
[18:19] the two tuned a super mode of 128 Ascend 950 chips. That's pretty
[18:24] 128 Ascend 950 chips. That's pretty
[18:24] 128 Ascend 950 chips. That's pretty interesting to see if you're not
[18:26] interesting to see if you're not
[18:26] interesting to see if you're not familiar. um uh Nvidia, the um AI chip
[18:32] familiar. um uh Nvidia, the um AI chip
[18:32] familiar. um uh Nvidia, the um AI chip maker, the the GPU
[18:35] maker, the the GPU
[18:35] maker, the the GPU um creator and kind of the hardware
[18:38] um creator and kind of the hardware
[18:38] um creator and kind of the hardware heavyweight that is leading uh or
[18:41] heavyweight that is leading uh or
[18:42] heavyweight that is leading uh or supporting a lot of this AI growth and
[18:45] supporting a lot of this AI growth and
[18:45] supporting a lot of this AI growth and buildout and model trainings and all of
[18:47] buildout and model trainings and all of
[18:47] buildout and model trainings and all of that. one of uh the the moes for Nvidia.
[18:52] that. one of uh the the moes for Nvidia.
[18:52] that. one of uh the the moes for Nvidia. I I think this is one of their moes. Um
[18:55] I I think this is one of their moes. Um
[18:55] I I think this is one of their moes. Um they have a software called CUDA. Um
[18:59] they have a software called CUDA. Um
[18:59] they have a software called CUDA. Um that sits on top of the chips and allows
[19:03] that sits on top of the chips and allows
[19:03] that sits on top of the chips and allows developers to program the chips
[19:06] developers to program the chips
[19:06] developers to program the chips specifically for the training runs and
[19:09] specifically for the training runs and
[19:09] specifically for the training runs and the use cases that u these different uh
[19:12] the use cases that u these different uh
[19:12] the use cases that u these different uh technologists have. Right. Um a lot of
[19:16] technologists have. Right. Um a lot of
[19:16] technologists have. Right. Um a lot of people have said this is um a a gamecher
[19:20] people have said this is um a a gamecher
[19:20] people have said this is um a a gamecher for Nvidia, right? Uh the the CUDA
[19:23] for Nvidia, right? Uh the the CUDA
[19:23] for Nvidia, right? Uh the the CUDA software because it enables so many
[19:26] software because it enables so many
[19:26] software because it enables so many developers to build on top of uh the
[19:29] developers to build on top of uh the
[19:29] developers to build on top of uh the CUDA chipset or or the Nvidia chipset,
[19:32] CUDA chipset or or the Nvidia chipset,
[19:32] CUDA chipset or or the Nvidia chipset, but also um it uh
[19:36] but also um it uh
[19:36] but also um it uh like once you get locked into it,
[19:38] like once you get locked into it,
[19:38] like once you get locked into it, there's there's nothing else that you
[19:40] there's there's nothing else that you
[19:40] there's there's nothing else that you can use. There's no open- source
[19:41] can use. There's no open- source
[19:41] can use. There's no open- source software that you can use on Nvidia
[19:43] software that you can use on Nvidia
[19:43] software that you can use on Nvidia chips essentially. So if you want to use
[19:44] chips essentially. So if you want to use
[19:44] chips essentially. So if you want to use Nvidia chips, you got to use CUDA
[19:46] Nvidia chips, you got to use CUDA
[19:46] Nvidia chips, you got to use CUDA software. China um and you know Huawei
[19:51] software. China um and you know Huawei
[19:51] software. China um and you know Huawei is trying to build chips to compete with
[19:53] is trying to build chips to compete with
[19:53] is trying to build chips to compete with Nvidia. Deepseek is coming out with a
[19:57] Nvidia. Deepseek is coming out with a
[19:57] Nvidia. Deepseek is coming out with a software that allows uh developers to
[20:02] software that allows uh developers to
[20:02] software that allows uh developers to you know program the chips to to train
[20:04] you know program the chips to to train
[20:04] you know program the chips to to train them uh train models and and run the
[20:06] them uh train models and and run the
[20:06] them uh train models and and run the chips how they want. Amazing. I mean
[20:08] chips how they want. Amazing. I mean
[20:08] chips how they want. Amazing. I mean this is pretty pretty interesting to see
[20:10] this is pretty pretty interesting to see
[20:10] this is pretty pretty interesting to see this coming out. Um, this is and
[20:13] this coming out. Um, this is and
[20:13] this coming out. Um, this is and especially from an open- source
[20:14] especially from an open- source
[20:14] especially from an open- source standpoint, uh, it's an under an MIT
[20:17] standpoint, uh, it's an under an MIT
[20:17] standpoint, uh, it's an under an MIT license, which means basically anybody
[20:19] license, which means basically anybody
[20:19] license, which means basically anybody can use it. Um, the Yeah, maybe the the
[20:24] can use it. Um, the Yeah, maybe the the
[20:24] can use it. Um, the Yeah, maybe the the catch here is, uh, there's still a
[20:26] catch here is, uh, there's still a
[20:26] catch here is, uh, there's still a supply issue. You know, Huawei says
[20:28] supply issue. You know, Huawei says
[20:28] supply issue. You know, Huawei says demand exceeds its capacity and fast
[20:33] demand exceeds its capacity and fast
[20:33] demand exceeds its capacity and fast single kernels don't prove full training
[20:36] single kernels don't prove full training
[20:36] single kernels don't prove full training runs, right? So
[20:39] runs, right? So
[20:39] runs, right? So this is an interesting story. You know,
[20:42] this is an interesting story. You know,
[20:42] this is an interesting story. You know, we continue to see new advancements come
[20:45] we continue to see new advancements come
[20:45] we continue to see new advancements come out of China. Um and they are uh
[20:48] out of China. Um and they are uh
[20:48] out of China. Um and they are uh formidable, right? Um definitely
[20:52] formidable, right? Um definitely
[20:52] formidable, right? Um definitely fighters, definitely super smart,
[20:53] fighters, definitely super smart,
[20:53] fighters, definitely super smart, working really really hard. um and and
[20:57] working really really hard. um and and
[20:57] working really really hard. um and and trying to push the envelope of their
[20:59] trying to push the envelope of their
[20:59] trying to push the envelope of their technology and what they can accomplish
[21:02] technology and what they can accomplish
[21:02] technology and what they can accomplish and um and and so hats off to Deep Seek
[21:05] and um and and so hats off to Deep Seek
[21:05] and um and and so hats off to Deep Seek and the Deep Seek team to be able to to
[21:07] and the Deep Seek team to be able to to
[21:07] and the Deep Seek team to be able to to to do this and open this up. It'll be
[21:09] to do this and open this up. It'll be
[21:09] to do this and open this up. It'll be interesting to see um how this impacts
[21:14] interesting to see um how this impacts
[21:14] interesting to see um how this impacts uh competitiveness and um how this
[21:18] uh competitiveness and um how this
[21:18] uh competitiveness and um how this impacts Nvidia, how this impacts the
[21:21] impacts Nvidia, how this impacts the
[21:21] impacts Nvidia, how this impacts the Frontier model companies. You know, time
[21:23] Frontier model companies. You know, time
[21:23] Frontier model companies. You know, time will tell on this, right? I think um I
[21:26] will tell on this, right? I think um I
[21:26] will tell on this, right? I think um I think we're going to continue to see
[21:28] think we're going to continue to see
[21:28] think we're going to continue to see Chinese labs um they're super scrappy,
[21:31] Chinese labs um they're super scrappy,
[21:31] Chinese labs um they're super scrappy, right? Like you know something like a
[21:34] right? Like you know something like a
[21:34] right? Like you know something like a tenth of the compute power of the United
[21:37] tenth of the compute power of the United
[21:37] tenth of the compute power of the United States is is what the Chinese have and
[21:39] States is is what the Chinese have and
[21:39] States is is what the Chinese have and and they're fighting, right? They're
[21:41] and they're fighting, right? They're
[21:41] and they're fighting, right? They're they're trying to figure out um the best
[21:43] they're trying to figure out um the best
[21:43] they're trying to figure out um the best the smartest ways, most efficient ways.
[21:45] the smartest ways, most efficient ways.
[21:45] the smartest ways, most efficient ways. Great thing is they're publishing most
[21:46] Great thing is they're publishing most
[21:46] Great thing is they're publishing most of most of this as open source uh which
[21:49] of most of this as open source uh which
[21:49] of most of this as open source uh which can add a lot to the overall uh AI
[21:52] can add a lot to the overall uh AI
[21:52] can add a lot to the overall uh AI ecosystem uh and how everybody builds.
[21:57] ecosystem uh and how everybody builds.
[21:57] ecosystem uh and how everybody builds. Uh, next up, um, Meta.
[22:01] Uh, next up, um, Meta.
[22:01] Uh, next up, um, Meta. The, uh, the New York Times reported
[22:03] The, uh, the New York Times reported
[22:03] The, uh, the New York Times reported Wednesday that Meta has been treating
[22:05] Wednesday that Meta has been treating
[22:05] Wednesday that Meta has been treating its AI data centers,
[22:08] its AI data centers,
[22:08] its AI data centers, um, as quote unquote pilot models, which
[22:11] um, as quote unquote pilot models, which
[22:11] um, as quote unquote pilot models, which lets it claim federal research tax
[22:14] lets it claim federal research tax
[22:14] lets it claim federal research tax credits on NVIDIA chips. Um, according
[22:18] credits on NVIDIA chips. Um, according
[22:18] credits on NVIDIA chips. Um, according to coverage of the report, Meta's
[22:20] to coverage of the report, Meta's
[22:20] to coverage of the report, Meta's research credit grew from $700 million
[22:23] research credit grew from $700 million
[22:23] research credit grew from $700 million in 2023 to $3.9 billion in 2025.
[22:29] in 2023 to $3.9 billion in 2025.
[22:29] in 2023 to $3.9 billion in 2025. Um,
[22:31] Um,
[22:31] Um, taxpayers might not like that a lot. Um,
[22:36] taxpayers might not like that a lot. Um,
[22:36] taxpayers might not like that a lot. Um, can you blame Meta for trying? Maybe
[22:38] can you blame Meta for trying? Maybe
[22:38] can you blame Meta for trying? Maybe not. I guess sounds like just a a
[22:40] not. I guess sounds like just a a
[22:40] not. I guess sounds like just a a loophole, but man, I I feel like Meta um
[22:45] loophole, but man, I I feel like Meta um
[22:45] loophole, but man, I I feel like Meta um shouldn't be playing loopholes right
[22:46] shouldn't be playing loopholes right
[22:46] shouldn't be playing loopholes right now. Meta says it's using incentives
[22:50] now. Meta says it's using incentives
[22:50] now. Meta says it's using incentives Congress created to encourage domestic
[22:52] Congress created to encourage domestic
[22:52] Congress created to encourage domestic research, investment, and employment. Um
[22:56] research, investment, and employment. Um
[22:56] research, investment, and employment. Um coverage of the report cites Meta's own
[22:58] coverage of the report cites Meta's own
[22:58] coverage of the report cites Meta's own 10K filings where the credits appear as
[23:01] 10K filings where the credits appear as
[23:01] 10K filings where the credits appear as a risk factor. You know, if I'm meta
[23:05] a risk factor. You know, if I'm meta
[23:05] a risk factor. You know, if I'm meta right now, I saw an interview yesterday,
[23:07] right now, I saw an interview yesterday,
[23:08] right now, I saw an interview yesterday, a clip where Zuck is talking about
[23:11] a clip where Zuck is talking about
[23:11] a clip where Zuck is talking about deploying $600 billion through 2028 on
[23:15] deploying $600 billion through 2028 on
[23:15] deploying $600 billion through 2028 on AI development and um buildout and um
[23:20] AI development and um buildout and um
[23:20] AI development and um buildout and um just mass I mean, can you just imagine
[23:23] just mass I mean, can you just imagine
[23:23] just mass I mean, can you just imagine $600 billion? and he had a really good
[23:25] $600 billion? and he had a really good
[23:25] $600 billion? and he had a really good point that um not spending that money
[23:28] point that um not spending that money
[23:28] point that um not spending that money and risking being behind or being late
[23:31] and risking being behind or being late
[23:31] and risking being behind or being late even if it means
[23:33] even if it means
[23:33] even if it means wasting a couple hundred billion
[23:35] wasting a couple hundred billion
[23:35] wasting a couple hundred billion dollars.
[23:37] dollars.
[23:37] dollars. Um it's still worth it, right?
[23:40] Um it's still worth it, right?
[23:40] Um it's still worth it, right? Uh from the outside looking in, man,
[23:43] Uh from the outside looking in, man,
[23:43] Uh from the outside looking in, man, that sounds crazy. Uh when you're inside
[23:46] that sounds crazy. Uh when you're inside
[23:46] that sounds crazy. Uh when you're inside this, that totally makes sense and and I
[23:47] this, that totally makes sense and and I
[23:47] this, that totally makes sense and and I agree with his point. That being said,
[23:51] agree with his point. That being said,
[23:51] agree with his point. That being said, you know, you got to work on your image
[23:52] you know, you got to work on your image
[23:52] you know, you got to work on your image a little bit. Um, of course, working
[23:55] a little bit. Um, of course, working
[23:55] a little bit. Um, of course, working within the letter of law, take advantage
[23:56] within the letter of law, take advantage
[23:56] within the letter of law, take advantage of it any way you can. But I, you know,
[23:59] of it any way you can. But I, you know,
[23:59] of it any way you can. But I, you know, it's interesting how, you know, I
[24:01] it's interesting how, you know, I
[24:01] it's interesting how, you know, I parallel this with what Gavin Newsome's
[24:03] parallel this with what Gavin Newsome's
[24:03] parallel this with what Gavin Newsome's doing in in California. And my point
[24:05] doing in in California. And my point
[24:05] doing in in California. And my point there is like, hey, look for the tax
[24:07] there is like, hey, look for the tax
[24:07] there is like, hey, look for the tax credits, but how can you pour that back
[24:09] credits, but how can you pour that back
[24:09] credits, but how can you pour that back into the community? Not that that's an
[24:11] into the community? Not that that's an
[24:11] into the community? Not that that's an obligation, but as just a goodwill
[24:13] obligation, but as just a goodwill
[24:13] obligation, but as just a goodwill thing. Like if you're looking at
[24:15] thing. Like if you're looking at
[24:15] thing. Like if you're looking at spending $600 billion in the next couple
[24:17] spending $600 billion in the next couple
[24:17] spending $600 billion in the next couple years, um if you are getting a tax
[24:20] years, um if you are getting a tax
[24:20] years, um if you are getting a tax break, why not pour that back into the
[24:23] break, why not pour that back into the
[24:23] break, why not pour that back into the local community, into education, into AI
[24:26] local community, into education, into AI
[24:26] local community, into education, into AI development, um a and spread some
[24:30] development, um a and spread some
[24:30] development, um a and spread some goodwill there? That's what I would love
[24:31] goodwill there? That's what I would love
[24:32] goodwill there? That's what I would love to see is just maybe a little bit more
[24:33] to see is just maybe a little bit more
[24:34] to see is just maybe a little bit more of that, a little bit more of
[24:35] of that, a little bit more of
[24:35] of that, a little bit more of enablement. Uh, and we need those good
[24:37] enablement. Uh, and we need those good
[24:38] enablement. Uh, and we need those good stories, too. Especially when, you know,
[24:40] stories, too. Especially when, you know,
[24:40] stories, too. Especially when, you know, we continue to see Open AI agent leaks
[24:42] we continue to see Open AI agent leaks
[24:42] we continue to see Open AI agent leaks and, uh,
[24:45] and, uh,
[24:45] and, uh, just, you know, fear-mongering and
[24:47] just, you know, fear-mongering and
[24:48] just, you know, fear-mongering and doommaxing over there. We'd love to see
[24:50] doommaxing over there. We'd love to see
[24:50] doommaxing over there. We'd love to see some goodwill from Meta. All in all, I
[24:53] some goodwill from Meta. All in all, I
[24:53] some goodwill from Meta. All in all, I think
[24:55] think
[24:55] think I think Meta's on a tear. I think uh
[24:57] I think Meta's on a tear. I think uh
[24:57] I think Meta's on a tear. I think uh Zuckerberg is um
[25:01] Zuckerberg is um
[25:01] Zuckerberg is um like he's a winner. You don't want to
[25:03] like he's a winner. You don't want to
[25:03] like he's a winner. You don't want to fight with him. Um, that being said,
[25:05] fight with him. Um, that being said,
[25:05] fight with him. Um, that being said, would love to see some more Goodwill
[25:07] would love to see some more Goodwill
[25:07] would love to see some more Goodwill coming out of that group over there. And
[25:10] coming out of that group over there. And
[25:10] coming out of that group over there. And then lastly, we got a kind of a fun
[25:13] then lastly, we got a kind of a fun
[25:13] then lastly, we got a kind of a fun story. Uh, Door Dash launch launched a
[25:16] story. Uh, Door Dash launch launched a
[25:16] story. Uh, Door Dash launch launched a product called Ask Door Dash on
[25:18] product called Ask Door Dash on
[25:18] product called Ask Door Dash on Wednesday. It's an AI agent. You text uh
[25:21] Wednesday. It's an AI agent. You text uh
[25:21] Wednesday. It's an AI agent. You text uh straight in iMessage and you say, "Hey,
[25:24] straight in iMessage and you say, "Hey,
[25:24] straight in iMessage and you say, "Hey, order my usual and it finds the food,
[25:26] order my usual and it finds the food,
[25:26] order my usual and it finds the food, builds the cart, checks out." Um, and it
[25:29] builds the cart, checks out." Um, and it
[25:29] builds the cart, checks out." Um, and it also unveiled its own delivery drone,
[25:31] also unveiled its own delivery drone,
[25:31] also unveiled its own delivery drone, Door Dash Air. um
[25:35] Door Dash Air. um
[25:35] Door Dash Air. um in uh tested it in Northern California
[25:38] in uh tested it in Northern California
[25:38] in uh tested it in Northern California at a Chipotle and a Popeye's. This is
[25:41] at a Chipotle and a Popeye's. This is
[25:41] at a Chipotle and a Popeye's. This is awesome. I love it. Um this is uh kind
[25:44] awesome. I love it. Um this is uh kind
[25:44] awesome. I love it. Um this is uh kind of really cool, kind of interesting. Um
[25:47] of really cool, kind of interesting. Um
[25:48] of really cool, kind of interesting. Um we're seeing a lot of these personal
[25:49] we're seeing a lot of these personal
[25:49] we're seeing a lot of these personal agents come out. We talked about OpenAI
[25:51] agents come out. We talked about OpenAI
[25:51] agents come out. We talked about OpenAI dots. Um, of course we've got Grockbot,
[25:54] dots. Um, of course we've got Grockbot,
[25:54] dots. Um, of course we've got Grockbot, we've got uh the Meta Muse Personal
[25:57] we've got uh the Meta Muse Personal
[25:58] we've got uh the Meta Muse Personal Agent, we've got uh Instinct, uh Open
[26:02] Agent, we've got uh Instinct, uh Open
[26:02] Agent, we've got uh Instinct, uh Open Claw, course all these coming out.
[26:03] Claw, course all these coming out.
[26:03] Claw, course all these coming out. Interesting to see uh an actual um
[26:08] Interesting to see uh an actual um
[26:08] Interesting to see uh an actual um uh an actual provider create their own
[26:12] uh an actual provider create their own
[26:12] uh an actual provider create their own personal agent. Um which it's I I
[26:15] personal agent. Um which it's I I
[26:15] personal agent. Um which it's I I wonder, you know, I'm sure we'll see
[26:17] wonder, you know, I'm sure we'll see
[26:17] wonder, you know, I'm sure we'll see more competition like this. of kind of
[26:20] more competition like this. of kind of
[26:20] more competition like this. of kind of these third-party personal agents that
[26:22] these third-party personal agents that
[26:22] these third-party personal agents that can order Door Dash for you and then the
[26:25] can order Door Dash for you and then the
[26:25] can order Door Dash for you and then the Door Dashes of the world creating their
[26:27] Door Dashes of the world creating their
[26:27] Door Dashes of the world creating their own personal agent. Pretty amazing for
[26:30] own personal agent. Pretty amazing for
[26:30] own personal agent. Pretty amazing for them to uh jump on that and to be the
[26:33] them to uh jump on that and to be the
[26:33] them to uh jump on that and to be the first ones out with that. I feel like
[26:35] first ones out with that. I feel like
[26:35] first ones out with that. I feel like Door Dash is a very um uh kind of on the
[26:38] Door Dash is a very um uh kind of on the
[26:38] Door Dash is a very um uh kind of on the edge company and uh exciting for uh you
[26:44] edge company and uh exciting for uh you
[26:44] edge company and uh exciting for uh you know food delivery and all the other
[26:46] know food delivery and all the other
[26:46] know food delivery and all the other delivery stuff that they do to see see
[26:48] delivery stuff that they do to see see
[26:48] delivery stuff that they do to see see what happens there. Um they confirmed
[26:51] what happens there. Um they confirmed
[26:51] what happens there. Um they confirmed that it's iOS only. There's a US wait
[26:54] that it's iOS only. There's a US wait
[26:54] that it's iOS only. There's a US wait list and uh right now it's just a
[26:56] list and uh right now it's just a
[26:56] list and uh right now it's just a Northern California um pilot uh you
[27:00] Northern California um pilot uh you
[27:00] Northern California um pilot uh you probably Bay Area I assume. Um the
[27:05] probably Bay Area I assume. Um the
[27:06] probably Bay Area I assume. Um the you know uh I love this. I love pushing
[27:09] you know uh I love this. I love pushing
[27:09] you know uh I love this. I love pushing it forward. Um um I you know I wonder if
[27:13] it forward. Um um I you know I wonder if
[27:13] it forward. Um um I you know I wonder if there is going to be competition and and
[27:15] there is going to be competition and and
[27:15] there is going to be competition and and who and surely there will like obviously
[27:17] who and surely there will like obviously
[27:17] who and surely there will like obviously their their agent is going to compete
[27:19] their their agent is going to compete
[27:19] their their agent is going to compete within Instinct or with a Grockbot or
[27:22] within Instinct or with a Grockbot or
[27:22] within Instinct or with a Grockbot or dots. Um the the interesting thing is um
[27:26] dots. Um the the interesting thing is um
[27:26] dots. Um the the interesting thing is um and maybe this is where this leads is um
[27:30] and maybe this is where this leads is um
[27:30] and maybe this is where this leads is um I've got my personal agent, my Grobot or
[27:33] I've got my personal agent, my Grobot or
[27:34] I've got my personal agent, my Grobot or um my Metamuse or or whatever it may be
[27:37] um my Metamuse or or whatever it may be
[27:37] um my Metamuse or or whatever it may be um that I use that actually talks to the
[27:42] um that I use that actually talks to the
[27:42] um that I use that actually talks to the Door Dash agent and puts my order in.
[27:44] Door Dash agent and puts my order in.
[27:44] Door Dash agent and puts my order in. And maybe that's what we see. I think
[27:47] And maybe that's what we see. I think
[27:47] And maybe that's what we see. I think we're going to see and there's been a
[27:49] we're going to see and there's been a
[27:49] we're going to see and there's been a talk around this for maybe 18 months or
[27:52] talk around this for maybe 18 months or
[27:52] talk around this for maybe 18 months or more about agent to agent communication,
[27:55] more about agent to agent communication,
[27:55] more about agent to agent communication, agent agent to agent protocol and maybe
[27:57] agent agent to agent protocol and maybe
[27:57] agent agent to agent protocol and maybe this is the form factor in which uh we
[28:00] this is the form factor in which uh we
[28:00] this is the form factor in which uh we see
[28:02] see
[28:02] see that stuff come together. It'll be
[28:03] that stuff come together. It'll be
[28:03] that stuff come together. It'll be interesting um to see where this goes.
[28:06] interesting um to see where this goes.
[28:06] interesting um to see where this goes. But congrats to Door Dash and uh
[28:09] But congrats to Door Dash and uh
[28:09] But congrats to Door Dash and uh exciting for uh the team over there and
[28:12] exciting for uh the team over there and
[28:12] exciting for uh the team over there and to see what happens. Hey
[28:16] to see what happens. Hey
[28:16] to see what happens. Hey We covered a lot today. It was an
[28:18] We covered a lot today. It was an
[28:18] We covered a lot today. It was an exciting show. Really fun. I think
[28:21] exciting show. Really fun. I think
[28:21] exciting show. Really fun. I think interesting to see um you know parallels
[28:25] interesting to see um you know parallels
[28:25] interesting to see um you know parallels between different entities, different
[28:27] between different entities, different
[28:27] between different entities, different governments. Um seeing how folks are
[28:29] governments. Um seeing how folks are
[28:29] governments. Um seeing how folks are reacting to the AI news across the
[28:31] reacting to the AI news across the
[28:32] reacting to the AI news across the board. Um if you have any uh questions
[28:36] board. Um if you have any uh questions
[28:36] board. Um if you have any uh questions or comments or or would like to connect
[28:38] or comments or or would like to connect
[28:38] or comments or or would like to connect with us, please go to taws.live.
[28:43] with us, please go to taws.live.
[28:43] with us, please go to taws.live. That's to.live. live. Um, you'll be able
[28:46] That's to.live. live. Um, you'll be able
[28:46] That's to.live. live. Um, you'll be able to sign up for our newsletter there as
[28:49] to sign up for our newsletter there as
[28:49] to sign up for our newsletter there as well, too. Um, and then if nothing else,
[28:52] well, too. Um, and then if nothing else,
[28:52] well, too. Um, and then if nothing else, my name is Austin Wilson. We'll see you
[28:54] my name is Austin Wilson. We'll see you
[28:54] my name is Austin Wilson. We'll see you tomorrow morning at 6:00 a.m.
[29:28] Heat.
[29:29] Heat. Heat.