# OpenAI Offers USA 5% Equity, Anthropic and Samsung Plan Chip

An episode of [AI Chat: AI News on Claude, ChatGPT, Gemini, Grok, Perplexity](https://invisiblenetwork.ai/podcast/1669990482), 2 Jul 2026, 13 min.

Audio: https://content.rss.com/episodes/365073/2960785/ai-chat-podcast/2026_07_02_23_08_08_76ff7b4c-7d4f-4000-a43e-fec0ba551cbd.mp3

Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! In this episode, we cover reports that OpenAI offered the U.S. government a 5% equity stake as part of a broader proposal to strengthen America’s AI leadership. We also look at Anthropic and Samsung’s reported plans to develop a custom AI chip and what that could mean for reducing reliance on Nvidia hardware. Show Links • Get the top 80+ AI Models for $8.99 at AI Box: https://aibox.ai • How I Grow and Scale My Business with AI: https://www.skool.com/aihustle • Get the AI Chat Daily Newsletter: https://www.aichatdaily.com/newsletter

## Transcript

[0:02] Welcome to the AI Chat Podcast. I'm your host,
[0:04] Jaden Schaefer. Every day I cover cutting -edge
[0:07] AI news and talk with the leaders behind it,
[0:09] breaking down what it means for your life and
[0:11] business. OpenAI is offering the United States
[0:17] a 5 % stake in a sovereign AI wealth fund, and
[0:20] Bernie Sanders is pushing for a 50 % tax. Cursor's
[0:24] $60 billion SpaceX deal is putting their OpenAI
[0:27] and Anthropic relationship into question, as
[0:30] that's getting a little bit more tricky now that
[0:32] they've been purchased by one of their biggest
[0:33] competitors. Anthropic is currently in talks
[0:35] with Samsung to create a custom AI chip, and
[0:38] Meta is putting their smart glasses behind a
[0:41] MetaOne premium subscription. Will people pay?
[0:44] That is the question. Microsoft has just committed
[0:46] $2 .5 billion to Frontier. This is a new AI deployment
[0:49] company. This is not new. We've seen the same
[0:51] thing from OpenAI and from Anthrop... If you
[0:54] want to use AI to grow and scale your business,
[0:57] I have a free school community called AI Hustle
[1:00] where we have over 400 members and we're all
[1:02] sharing how we're growing and scaling our business
[1:04] using AI tools. My friend Jamie McCulley and
[1:07] I, we post a video every single week on there
[1:09] where we break down how we're using it inside
[1:11] of our own businesses. We share revenue, we share
[1:13] metrics, we share numbers and all of the software
[1:15] we use along the way. If you want to get access
[1:17] to over 150 videos, we have a premium section
[1:20] for $20 a month, but you can get free videos
[1:23] for free and access to the school community and
[1:25] our weekly live streams by going to school .com
[1:28] slash AI hustle. I'll leave a link in the description
[1:30] to help you grow and scale your business using
[1:33] AI tools. OpeningEye is offering to give the
[1:35] United States government, and I think really
[1:37] kind of the entire United States theoretically,
[1:39] a 5 % equity stake. This is kind of the same
[1:42] as Alaska. They have an oil wealth division fund.
[1:45] This is parts of some earlier talks that the
[1:48] Trump administration was doing. And right now,
[1:50] this is also sitting right between... big public
[1:53] pressure campaigns we're seeing. We have Senator
[1:55] Bernie Sanders that is, you know, he has a competing
[1:58] plan. He says he would like 50 % a one -time
[2:00] tax on all of the leading AI firms where they
[2:03] just give 50%. And so I think OpenAI is like
[2:06] happy to offer 5 % when the Bernie Sanders 50
[2:09] % is the other offer being proposed. Bernie Sanders
[2:12] proposal would raise about $7 trillion through
[2:15] a one -time 50 % stock tax, which is 50 % of
[2:18] the company and the proceeds and direct payments.
[2:21] I would go to healthcare and housing and other,
[2:23] you know. programs that Bernie Sanders deems
[2:26] fit. The Trump administration has separately
[2:29] approached Google and Meta about a very similar
[2:31] equity stake. I think neither of those companies
[2:33] has publicly said that they are going to sign
[2:35] any sort of agreement like that. And I think
[2:37] 70 % of Americans right now oppose AI data centers
[2:40] in the area. 50 % more have concern about excitement
[2:44] rather than excitement about AI. So if you look
[2:46] at some of the polling that's going on inside
[2:48] of the country, there is, you know, a lot of
[2:50] negative sentiment that goes towards AI. And
[2:52] I think that's what OpenAI and maybe the Trump
[2:54] administration is trying to help. Bernie Sanders
[2:56] wants to just take 50 % of it. We'll see which
[2:59] of these two models plays out in the end. Cursor
[3:02] was acquired for $60 billion by SpaceX, but,
[3:05] you know, famously, its two biggest, I guess,
[3:08] suppliers or customers or however you want to
[3:10] call it, are OpenAI and Anthropic, right? Cursor
[3:13] uses OpenAI and Anthropic to power their tools
[3:15] or to let developers do that. And so people are
[3:17] calling into question if this relationship can
[3:19] continue, right? If Cursor was purchased by SpaceX,
[3:22] which is going to have XAI and XAI code running
[3:25] things inside of Cursor. Can they keep a good
[3:28] relationship with OpenAI and Anthropic? We've
[3:30] seen other companies that have grown and been
[3:32] bought out go through very similar challenging
[3:35] relationships. Scale .ai was purchased, or 50
[3:37] % of it was purchased by Meta, and their CEO
[3:39] was brought on to run Meta .ai. And when that
[3:42] happened, some of their biggest customers were
[3:44] Google and OpenAI, and both of those companies
[3:46] cut ties with them because of the competition.
[3:49] Cursor is already training their own next model
[3:51] on SpaceX infrastructure. They have 10 to 20
[3:53] times more compute than they previously had.
[3:55] because SpaceX gave it to them. So obviously
[3:57] Elon Musk is investing in building their own
[4:00] native capabilities. But the precedent for these
[4:02] coding tools specifically, I think is stronger
[4:04] than just what happened to scale AI because Windsurf,
[4:07] another very famous coding tool, was purchased
[4:10] by OpenAI last year. And when that happened,
[4:12] Anthropic cut off their cloud access. So I think
[4:15] there's a lot of precedent. It's very possible
[4:18] that Cursor is going to get their OpenAI or cloud
[4:20] access cut off. But this one seems a little bit
[4:22] trickier because Cursor really is just synonymous
[4:24] with coding. in the industry. I mean, outside
[4:26] of Cloud Code and maybe Codex, like this is the
[4:28] biggest coding tool that exists out there. I
[4:31] would be, you know, shocked if it, you know,
[4:33] if they really did cut it off, because that would
[4:34] be a ton of users that no longer would be, would
[4:37] be, you know, essentially giving free money to
[4:40] OpenAI and Anthropic, especially because Cursor
[4:42] signed up with a lot of enterprise contracts.
[4:44] And actually one really funny turn of events
[4:46] in all of this is that OpenAI startup fund, they
[4:48] were actually an early investor in Cursor, and
[4:50] they're going to convert their stake into SpaceX
[4:53] stock. So with all of this acquisition that's
[4:55] going on, Elon Musk and Sam Altman, who both
[4:58] have a lot of beef, OpenAI is going to get a
[5:00] direct financial stake into Cursor's success
[5:03] and also XAI, even though Codex is a direct competitor.
[5:08] So that's kind of funny. So there is a big move
[5:11] happening with Anthropic right now. They're in
[5:13] talks with Samsung to create a custom AI chip.
[5:16] They're joining OpenAI, Google, Amazon, all of
[5:19] them that are trying to basically reduce the
[5:20] reliance on NVIDIA as NVIDIA is, you know, know,
[5:23] making record sales. Right now doing all of this
[5:26] is both kind of targeted at costs and supply
[5:28] constraints. When you make this custom silicone,
[5:31] it's in, you know, when these labs are creating
[5:33] it themselves, it's going to let them optimize
[5:35] their own architectures. And also they can negotiate
[5:38] leverage against any sort of dominant supplier,
[5:41] right? So NVIDIA has been very dominant for a
[5:43] long time. And in fact, when you're trying to
[5:45] go and get GPUs or whatnot from NVIDIA, they
[5:48] have a certain amount that they'll allocate to
[5:50] different companies, but they're trying to spread
[5:52] it out. because the demand is just so high. So
[5:55] this is not great because if someone like OpenAI,
[5:57] Anthropic, Google, they only can get a maximum
[5:59] amount from NVIDIA. So if they can build their
[6:01] own, not only does that kind of ease that constraint,
[6:04] that supply constraint, but also, you know, maybe
[6:06] they have some better negotiating power. Look,
[6:08] we can build our own for X price if you would
[6:11] be able to bring the price of yours down. So
[6:13] I think it's kind of a long -term play for negotiating
[6:15] power with NVIDIA as well. OpenAI also just unveiled
[6:17] their own custom chip. It's called a Jalapeno.
[6:20] They did this last week. that with Broadcom,
[6:22] and they're claiming that it's going to offer
[6:24] better performance per watt than any of the other
[6:26] competitors. And I think when you're, you know,
[6:28] when you're building this custom for your software
[6:30] or hardware stack, it's, you know, it makes a
[6:33] lot of sense that they're able to achieve those
[6:34] kind of efficiencies. So I think Anthropic has
[6:37] seen this and trying to do the same thing. Samsung
[6:40] already manufactures NVIDIA's chips, and they
[6:42] do that for their AI training. And they're also
[6:44] building a joint AI chip factory in South Korea
[6:47] with NVIDIA right now. So, you know, I think
[6:49] they're kind of at the center of the supply.
[6:51] chain Samsung's working with Nvidia. So it's
[6:53] kind of a no brainer for Anthropic to go there
[6:55] as a partner to build out this chip. Anthropic's
[6:58] project, though, is definitely in the early stages,
[7:00] these chips are not quick, they take a very long
[7:02] time to build. And the purpose, the server role,
[7:05] the performance targets, none of that have actually
[7:07] been finalized. So I think you can kind of imagine
[7:09] this is gonna be a multi year development timeline.
[7:11] I mean, I've been hearing, you know, rumors of
[7:13] opening eye working on this stuff for many years.
[7:16] And it seems like Anthropic is just getting there
[7:18] and realizing this is something that they need
[7:19] to do. In other news, Meta is putting their smart
[7:22] glass features behind a premium subscription.
[7:26] It's kind of interesting, right? Facebook has
[7:27] kind of always been known as the free forever.
[7:30] They just monetize through ads. They're starting
[7:33] to spin out these new products, and they're trying
[7:35] to do premium subscription models. So if you
[7:38] want to get all of the latest features with their
[7:40] smart glasses, you're going to have to pay a
[7:41] subscription that's starting with conversation
[7:44] focus, which is basically going to enhance voice
[7:46] clarity in noisy settings. So I've heard this
[7:49] kind of pitched as like if you're in a noisy
[7:50] train station, and there's a ton of people talking.
[7:53] If you're talking to someone, their voice is
[7:56] going to get isolated and amplified by the speakers
[7:58] in your headset, so you can actually hear them
[8:00] better. All you have to do is look in their direction.
[8:02] and you'll be able to hear the sound of their
[8:04] voice louder than all of the noisy surroundings.
[8:07] But what's interesting, they do give you some
[8:08] free. They give you three hours a month for free
[8:12] and they give you 15 hours for $10 a month of
[8:15] Meta 1 Premium. I'm not sure if you turn this
[8:17] on or if it automatically detects and tells you
[8:19] when it runs out. I mean, it's basically usage
[8:21] -based AI like every other platform. It's just
[8:23] interesting how they're going to try to make
[8:25] this seamless and if they're going to just, you
[8:28] know, all of a sudden you're halfway through
[8:29] a conversation, everything's being perfectly
[8:30] amplified. and then it cuts out, right? So, I
[8:32] mean, and I don't know like how many conversations
[8:34] if you're having more than 15 hours of conversations
[8:36] in noisy train stations or maybe on like a noisy
[8:39] job site, it'll be interesting to see where this
[8:41] gets rolled out. This is definitely a big monetization
[8:44] play. It's not kind of a cost recovery one. The
[8:47] feature runs entirely on device. So there's no
[8:49] server cost to meta. So they're just trying to
[8:51] generate revenue from this. I think this shows
[8:53] that AI hardware is gonna get a lot of value
[8:56] that they can extract after their users already
[8:59] have the glasses, which. means they can basically
[9:01] sell their glasses at a loss assuming they have
[9:04] you know this conversation focus and probably
[9:06] a bunch of other ai features they'll roll out
[9:08] in the future especially if it can be computed
[9:09] on device they're not incurring any extra cost
[9:12] for it you just pay for it honestly kind of reminds
[9:14] me of tesla and their self -driving how like
[9:17] the technology is already in the tesla cars and
[9:19] if you just pay for it um then all of a sudden
[9:21] it's enabled meta's new $299 meta branded glasses
[9:26] are sold at break even. So they're really trying
[9:29] to grow their adoption. You know, previously
[9:31] it was about $100 more if you wanted the meta
[9:34] Ray -Bans, but now they're just selling them
[9:35] by themselves. And there is a very razor thin
[9:39] model of, you know, monetization on this. And
[9:43] then basically the hardware gets cost and they're
[9:45] trying to get the reoccurring software subscriptions
[9:47] to drive all of the revenue. And they're not
[9:50] the only people doing this because Google is
[9:51] also launching a competing. smart glasses this
[9:54] year they're partnering with samson to do that
[9:56] and warby parker and gentle monster google's
[9:58] own inference efficiency gains i think are going
[10:01] to let them bundle a lot more features than meta
[10:03] is and they're not going to have as much of aggressive
[10:05] tiering i don't think but they're not the only
[10:08] ones doing this apple is also having a bunch
[10:11] of their AI features behind a paywall and Google
[10:13] has other ones. Google specifically with their
[10:15] pixel video boost, it's going to require a specific
[10:17] Google one tier on their pixel devices. And Gemini
[10:21] live is also behind the Gemini premium. Um, and
[10:24] if you even like with iOS 27 for Apple, they
[10:27] have a bunch of photo editing that is locked
[10:29] in higher iCloud plus plan. So if you want some
[10:32] of the better features, you're going to have
[10:33] to also pay for that. Microsoft is launching
[10:36] a new AI deployment company called frontier.
[10:38] They're putting 2 .5 billion. dollars into this
[10:41] it's going to be its very own unit and they're
[10:42] going to have about 6 000 engineers that are
[10:45] kind of be going to be directly embedded into
[10:48] different companies. And they're trying to basically
[10:50] deploy AI systems end to end. So this is something
[10:53] that we've seen OpenAI also spin out like separate
[10:56] companies and get investors in these separate
[10:58] deployment companies. Anthropic has also done
[11:01] this and they're partnering with a bunch of big
[11:03] players. So it's interesting. It feels like the
[11:05] next step for these labs as they're trying to
[11:07] get more sales and more users in companies is
[11:10] to just basically get these engineers and they're,
[11:12] you know, sending them in there for at cost.
[11:14] or free or charging them something. And they're
[11:17] going to go and implement all of the Microsoft
[11:19] or all of the OpenAI or all of the Anthropic
[11:20] systems inside of that company and try to get
[11:22] their usage up as high as possible. There is
[11:25] a huge development deployment race right now
[11:28] with AWS. They announced a rival $1 billion venture
[11:31] just two days earlier. And I think both of these
[11:33] companies are betting that enterprises need hands
[11:36] -on implementation help. So it's not just about,
[11:38] you know, hey, go sign up for OpenAI and let
[11:40] your team use it. It's like, look, we're going
[11:42] to give you these engineers that are going to...
[11:44] to go and actually implement this for you. Just
[11:46] to put that into context, though, with their
[11:47] 6 ,000 engineers, Frontier is about six times
[11:50] larger than Palantir's entire FTE organization
[11:53] when they first IPO'd. And it's about double.
[11:56] AWS's announced commitment. So, I mean, this
[11:58] is a huge move for them. They have a bunch of
[12:00] early partners, including the London Stock Exchange
[12:02] Group, Unilever, Lando Lakes, and Accenture.
[12:06] They have a mix of some of the big data -heavy
[12:08] supply chain system integrator customers. I mean,
[12:11] Microsoft's massive, so this is not a shocker
[12:13] in any way. But their existing Azure and Microsoft
[12:15] 365 relationships, I think, give them a lot of
[12:18] embedded engineers already inside of most of
[12:20] the Fortune 500 companies. This is a big structural
[12:23] advantage they have over AWS. which is going
[12:25] to have to build this from the ground up. So
[12:27] yeah, Microsoft has kind of a leg in this from
[12:30] everything they've been doing with Azure. If
[12:32] you guys want to get access to the top 80 different
[12:34] AI models for just $8 a month, go check out AI
[12:36] box .ai. I give you access to over 80 different
[12:40] AI models, everything from text, image, audio,
[12:43] video, even music generation models. You get,
[12:45] I mean, it's basically all the top models, opening
[12:46] eye, Google, Anthropic, 11 labs, Claude, all
[12:49] of that. And it's only $8 .99 a month. So if
[12:52] you want to get access to all of that in one
[12:53] platform for one subscription, go check out my
[12:56] own startup, AIbox .ai. I'll leave a link in
[12:58] the description, and I hope you all have a fantastic
[13:00] rest of your day.

Shows are listed from Apple Podcasts' charts and directory, and Podcast Index where it is used. Every show and episode belongs to its publisher and plays from the publisher's own feed; this site hosts none of them.