Latest / In the Loop AI / Anthropic Just Got Hacked by China. These are the New Front Lines.

Transcript

Anthropic Just Got Hacked by China. These are the New Front Lines. · In the Loop AI

0:00
Ejaaz: China just got exposed for stealing our AI. In a new report from Anthropic,
0:05
Ejaaz: three top Chinese AI labs were exposed for having 16 million fraudulent conversations
0:10
Ejaaz: with Claude with one specific goal, to try and steal its capabilities to train their own models.
0:16
Ejaaz: Now, the week before, Google said the same thing about China attacking their Gemini models.
0:20
Ejaaz: The week before that, OpenEye said the same thing. The top three American AI
0:24
Ejaaz: labs are blaming China for trying to hack their own AI models.
0:28
Ejaaz: But here's the twist in the story. What China's actually doing may not actually
0:32
Ejaaz: be illegal in the first place.
0:34
Ejaaz: In fact, this is something that every AI company is doing to get ahead in the AI race.
0:39
Ejaaz: In this episode, we're going to explore what all these reports confirm and whether
0:42
Ejaaz: distillation, the hacking vector, is actually a bad thing.
0:46
Josh: Yeah, so it starts with this blog post that Anthropik published earlier this
0:48
Josh: week that says, it's a title, Detecting and Preventing Distillation Attacks.
0:52
Josh: And I guess maybe it's helpful to just kind of define distillation as a concept
0:56
Josh: before we get into what they're accusing China of.
0:58
Josh: And basically, the way it works is there is a teacher and a student model.
1:02
Josh: So the teacher is the large model. That would be Anthropik's Claude Opus model. It's this huge model.
1:07
Josh: They've spent hundreds of millions, billions of dollars training it and turning
1:10
Josh: it into the model that we use every day.
1:12
Josh: That model provides these high quality outputs to the student,
1:16
Josh: which is the smaller model that is getting distilled from.
1:19
Josh: So basically the smaller model, the distilled model learns to mimic the outputs
1:23
Josh: of the larger model, but does so at a fraction of the cost because it's able
1:26
Josh: to kind of cherry pick the types of outputs that it gets by prompting it very specifically.
1:31
Josh: So anybody with sufficient access to a model and enough prompts can actually
1:35
Josh: get enough information to emulate the large model with a much smaller data set.
1:39
Josh: Now the outputs are not always as good as the large model, but they're significantly
1:43
Josh: cheaper and oftentimes very close.
1:45
Josh: So in the case that you get an extra breakthrough or two on top of that,
1:48
Josh: you can build a pretty impressive model.
1:50
Josh: And allegedly, that's what's happening with these models from China,
1:53
Josh: at least according to Anthropic.
1:54
Ejaaz: At least that's what they say. But why is what you just described a good thing?
1:59
Ejaaz: It's because all the hundreds of billions of dollars that are invested in building
2:03
Ejaaz: out the best AI model isn't sustainable for the long-term future.
2:08
Ejaaz: In fact, if you want to have a model that's small enough to fit on your phone,
2:11
Ejaaz: but as intelligent enough as the top models, it needs to be distilled through
2:15
Ejaaz: that process that you just explained.
2:17
Ejaaz: So it's going from a big model to a smaller model that is just as intelligent
2:20
Ejaaz: in certain specific ways.
2:22
Ejaaz: The stats from this China hack on Anthropic, Josh, are kind of insane.
2:27
Ejaaz: So I mentioned 16 million exchanges, but they spun up 24,000 fake Anthropic
2:32
Ejaaz: accounts. Now, I have to specify it.
2:35
Ejaaz: Anthropic does not allow Chinese users to access their models for the specific
2:40
Ejaaz: reason that adversaries to the U.S.
2:43
Ejaaz: Could get access to superintelligence that they're building.
2:46
Ejaaz: So DeepSeek, I'm going to name some names now. DeepSeek, one of the top AI labs
2:50
Ejaaz: which caused the stock market to crash at the end of 2024, I believe,
2:54
Ejaaz: were responsible for 150,000 of those exchanges. Moonshot AI, 3.4 million.
2:59
Ejaaz: Minimax, which is a favorite that you and I have spoken about on the show,
3:03
Ejaaz: 13 million exchanges of those 15-minute conversations.
3:07
Ejaaz: There is an argument here that the open source gold rush that has been happening
3:11
Ejaaz: in China was mainly because they were stealing US secrets.
3:14
Josh: Wouldn't that be funny if that was the case? And then if that's also the case,
3:17
Josh: then what do you do about it? I mean, Anthropic kind of came out and they were
3:20
Josh: very upset about this, clearly.
3:22
Josh: But at the end of the day, it's like kind of on them. the onus is on them to
3:26
Josh: protect their systems and prevent this from happening there's a really great
3:30
Josh: great post that you have on screen and it's a joke it says my
3:33
Josh: son asking me a lot of questions it's a distillation attack obviously
3:36
Josh: and i think it's it's kind of funny where like the irony is and we can get into
3:40
Josh: the hypocrisy of the whole thing is that anthropic as a company very much has
3:45
Josh: done this in the past in order to get where they are and they are kind of the
3:49
Josh: person who's crying wolf now saying wait we're getting attacked this is not
3:53
Josh: allowed we should not be able to do this.
3:55
Ejaaz: Yeah, I mean, Anthropic is doing this with their own models,
3:58
Ejaaz: right? They've distilled Claude Opus into their Haiku model.
4:01
Ejaaz: Google's distilled Gemini Ultra into Gemini Nano. This is a common practice.
4:05
Ejaaz: So then the question becomes, which part of this is illegal?
4:09
Ejaaz: What has China done that is illegal?
4:10
Ejaaz: Well, it's two things that Anthropic has claimed. Number one,
4:13
Ejaaz: they've got this fancy terms of service, which they're lawyers that have been
4:16
Ejaaz: paid millions of dollars that have drafted up, which says, hey,
4:19
Ejaaz: hey, hey, if it's our model that you're doing this to,
4:22
Ejaaz: you can't do that. We've patented this thing. It's going to be illegal and we're
4:25
Ejaaz: going to sue you in a court of law.
4:27
Ejaaz: The issue is China is in China and they don't abide to the US legal system at all.
4:32
Ejaaz: Which brings me to the second thing that they violated, which is a geographical
4:35
Ejaaz: restriction. They don't let anyone in that region access clause.
4:39
Ejaaz: So the fact that China has been able to pull this off from the top AI labs means
4:43
Ejaaz: that they've illegally spun up accounts to do this.
4:45
Josh: Well, you know who doesn't care about laws is China.
4:49
Josh: Like they could not care less. In fact, this is the time for wartime CEOs.
4:55
Josh: Like in very many ways, this is the largest war that's being fought between
5:00
Josh: the US and China. And it's around AI.
5:02
Josh: And I think for them to say, that's against our terms of service, this is wrong.
5:06
Josh: Like that is not a grounds for defending yourself because clearly they have
5:11
Josh: no disregard, they have no regard for any sort of law.
5:14
Josh: I mean, you'd look at Seed Dance 2.0 and how it violates every copyright law under the sun.
5:18
Josh: And yet people don't care. It's the best video generation model in the world that exists.
5:23
Josh: So it is a challenge to claim that because they're violating terms of service,
5:27
Josh: this is an illegal thing that you shouldn't be able to do.
5:29
Josh: And they've not just cut off China, but I think it's important to note they've
5:33
Josh: also cut off other frontier AI labs.
5:35
Josh: They famously had this beef with XAI recently where they cut off all of the
5:39
Josh: Claude Code access to other labs.
5:41
Josh: So Anthropica has been very controlled and closed down in who's actually able
5:45
Josh: to access their models. And it sounds like someone was able to bypass that and
5:48
Josh: they just got pretty upset about it.
5:51
Ejaaz: Well, I mean, the Pentagon is relying on the likes of Anthropic,
5:56
Ejaaz: XAI, and OpenAI to fund the warfare effort against China, right?
6:00
Ejaaz: To your point, we're in like a wartime position. Like these AI models are being
6:03
Ejaaz: used as a geopolitical weapon.
6:05
Ejaaz: And so whoever owns the best model per se can advance the quickest.
6:10
Ejaaz: So it's like an economically dependent thing. And this whole drama with the
6:14
Ejaaz: Pentagon has been, the Pentagon has been using Claude for pretty much quite
6:19
Ejaaz: a lot of covert activity, including the recent capture of Nicolas Maduro,
6:23
Ejaaz: the former, I guess, president of Venezuela.
6:26
Ejaaz: And the issue now is that Anthropic is restricting Pentagon's access,
6:31
Ejaaz: like American owned self-defense against these kinds of things.
6:35
Ejaaz: And so the Pentagon is getting fed up and issuing them an ultimatum and saying,
6:37
Ejaaz: listen, if you don't figure this out, we're going to classify you as a threat to the country.
6:42
Ejaaz: Now, I have to give credit to Anthropic for maintaining their identity evenly
6:48
Ejaaz: across every single facet, but
6:49
Ejaaz: I don't think it's the smart way to do it because at the end of the day,
6:53
Ejaaz: there are going to be things that require more uncensored versions and you just
6:55
Ejaaz: need to be compliant with that fact.
6:57
Ejaaz: Because to your point earlier, Josh, Claude, OpenAI, ChatGPT has become a national
7:03
Ejaaz: asset. And so it needs to be treated as such.
7:06
Josh: Yeah, it's a matter of national security. And the thing about Anthropic that's
7:08
Josh: unique to Anthropic, and I'm not sure many other companies in the AI space is
7:13
Josh: their mission statement, where if you talk to any employee who works at Anthropic,
7:16
Josh: they'll tell you the purpose of the company is safety and alignment.
7:20
Josh: And I think while it's a valiant effort and incredibly important,
7:25
Josh: it doesn't really bode well for the current state of affairs in which
7:28
Josh: velocity momentum and just raw speed to get
7:31
Josh: to the best model possible is actually beneficial so
7:35
Josh: i think what we're seeing here is there's just these increasing conflicts
7:38
Josh: with i mean the secretary of defense and the pentagon wanting
7:41
Josh: access to do things that they deem to
7:44
Josh: be a matter of national security and like xai wanting to go and build code using
7:49
Josh: their tools they're like no no no no that's not how we want this used we're
7:52
Josh: not going to allow that and then the rumor is is that apparently the pentagon
7:55
Josh: actually just kicked out anthropic and now grok and the xai team is responsible
7:59
Josh: for being the AI provider for the Pentagon.
8:02
Josh: So I found that interesting too. It's just like a little side development.
8:05
Ejaaz: Well, I mean, like what you're getting at there is that some of these AI models
8:09
Ejaaz: or AI companies in America are kind of being super hypocritical.
8:13
Ejaaz: Like this tweet actually explains it really well.
8:16
Ejaaz: Hey, did you hear about the little like $1.5 billion lawsuit that Anthropic
8:22
Ejaaz: had to pay out over pirating or illegally downloading 7 million books to train their own models?
8:28
Ejaaz: Open AI is facing similar lawsuits against newspapers or across newspapers,
8:32
Ejaaz: code repositories and authors.
8:34
Ejaaz: I'm pretty sure Anthropic got sued for using Reddit data to train their models.
8:37
Ejaaz: Google trained their entire model over the index data that they took.
8:41
Ejaaz: Now, the question then becomes, is that fair?
8:43
Ejaaz: Who are paying the authors and creators of the content where these AI labs that
8:48
Ejaaz: have like amassed hundreds of billions of dollars worth of valuation,
8:51
Ejaaz: who's paying those creators?
8:53
Ejaaz: No one is, right? So you could argue that that is a form of distillation.
8:56
Ejaaz: Now, obviously, that's looking at it in a very black and white face,
8:59
Ejaaz: but I do think it's hypocritical.
9:01
Ejaaz: And most importantly, the memes are just so, so good here.
9:05
Ejaaz: You've got people that are asking Claude in Chinese, what model are you?
9:08
Ejaaz: And them replying, hey, I'm DeepSeek.
9:11
Ejaaz: And then you've got this one here where it says, I can't believe someone would
9:13
Ejaaz: just steal from Anthropic like this.
9:15
Ejaaz: Anthropic spent millions of man hours handwriting code, text,
9:19
Ejaaz: art, and books. Obviously,
9:20
Ejaaz: you know tongue-in-cheek this isn't actually real the point that's being made
9:23
Ejaaz: is that all information is kind of taken or stolen or interpreted in some way
9:28
Ejaaz: shape or form so what makes it any different for china in this regard.
9:31
Josh: The crux the argument is that the same foundation
9:35
Josh: that anthropic built its models on is the foundation that chinese models are
9:39
Josh: building their foundation on it's just one level kind of up where they clearly
9:44
Josh: stole maybe not stole content but they clearly used the content that we've produced
9:49
Josh: as humans over time to train their model.
9:52
Josh: What Deep Seek is doing is the next layer up. It's taking the,
9:55
Josh: I guess, the quantized version of all of the human intelligence that we've developed
9:58
Josh: and then distilling that one layer up.
9:59
Josh: It's easy to see why they would be upset, but it's also easy to see why everyone
10:04
Josh: is kind of deeming them as hypocritical.
10:06
Josh: It's like, again, you know that you are a nation state actor,
10:10
Josh: like relatives to the rest of the world in one of the most important wars that's
10:12
Josh: being fought. You know that you are going to be getting attacked.
10:15
Josh: You know that these people are going to be coming for you to build their own
10:17
Josh: models in the race for this AGI and beyond.
10:20
Josh: And to think that it's not going to happen and to be upset when it does just seems wrong.
10:27
Josh: And I think that's probably where a lot of the backlash is coming from is because,
10:31
Josh: I mean, again, it's on them to solve for these issues before they happen or
10:35
Josh: accept the consequences if they don't.
10:37
Josh: And that's just what happens here. I mean, there's like, this is a bar fight.
10:42
Josh: There are no rules in this fight. It is the only thing that you're trying to
10:44
Josh: do is get to AGI as fast as possible. And clearly, China doesn't care.
10:49
Ejaaz: Can I say something in China's defense? And maybe this is a hot take.
10:53
Ejaaz: Their models be banging recently, okay?
10:56
Ejaaz: Like they have been churning out new model updates from the likes of Alibaba with Quen 3.5.
11:02
Ejaaz: By the way, if you haven't tried this model out, apparently it's really amazing
11:05
Ejaaz: with agents. It's absolutely crushed benchmarks. Once again, open source.
11:09
Ejaaz: We've got Minimax AI that we mentioned earlier, which was the biggest perpetrator
11:12
Ejaaz: of this distillation attack against Anthropic.
11:15
Ejaaz: It's the most used model on Open Router.
11:17
Josh: Also, what's interesting is like Minimax 2.5 is the most popular Chinese model for Open Claw too.
11:22
Josh: And i personally used it like when i was running into
11:25
Josh: the oauth issues with claude because they were they were kind
11:28
Josh: of threatening again they were threatening to ban users for using oauth for
11:30
Josh: going around things they're just the hardos like they have no fun uh but when
11:34
Josh: they were threatening to like break people's accounts and ban them i switched
11:38
Josh: over to minimax 2.5 and it actually worked very well and it's a fraction of
11:42
Josh: the cost and i was like hey if you're gonna push me away i'm gonna go here to
11:46
Josh: these models that get the job done for me and minimax was that one I
11:49
Ejaaz: Have a question for you like where are you geographically located right now are you in China.
11:54
Josh: No, I'm certainly not.
11:55
Ejaaz: Okay, so it looks like they're just giving you free access to do these things.
11:58
Ejaaz: There's no geographical jurisdictions that they're kind of like placing on your
12:02
Ejaaz: restrictions. They're just letting you do the thing.
12:04
Ejaaz: It's awesome. Like all of these models are open source.
12:07
Ejaaz: These are kind of embellishments that America should be propagating,
12:11
Ejaaz: but they're not. They're playing the opposite.
12:13
Ejaaz: They're playing kind of secretive and it's not working out in their favor.
12:16
Ejaaz: I mean, you've got Minimax, all these latest Chinese labs, by the way, GLM-5,
12:20
Ejaaz: Kimi-K 2.5, minimax are crazy good
12:23
Ejaaz: at computer use and agentic tooling kimike 2.5
12:27
Ejaaz: actually for the open claw fans out there released a
12:30
Ejaaz: browser extension and it's actually really good because the
12:32
Ejaaz: major issue with using open claw was that there was security issues
12:35
Ejaaz: well they created a sandbox environment that you can now use it so they're innovating
12:39
Ejaaz: at scale and to the case that they might be stealing certain secrets i don't
12:43
Ejaaz: think this is regarded as a hack or a stealing thing i actually just think they're
12:47
Ejaaz: trying to get better models out to more people and hey if america can use it
12:50
Ejaaz: it's hardly a geopolitical thing so i don't know i'm kind of in the defense
12:54
Ejaaz: of china here and maybe that's a hot take and then kind of finally i just want to one.
12:58
Josh: Thing on that china note i'm not sure that it's out of their own goodwill of
13:03
Josh: their heart i think it's like the reason they're open source is probably because
13:07
Josh: they're behind i have i would imagine that if they did have an nvidia equivalent
13:12
Josh: in China that was creating top tier GPUs and
13:16
Ejaaz: They did have the.
13:17
Josh: Yeah they would did have these leading models like Opus 4.6 and
13:20
Josh: GPT 5.3 they would close it
13:23
Josh: down because there is so much value in owning that but because China's behind
13:27
Josh: there's value in being open and sharing it and gaining as much adoption as possible
13:31
Josh: as quickly as possible and it seems like it's more strategic and tactical than
13:36
Josh: out of the the goodness of their heart but yeah I mean again open source really benefits everyone.
13:41
Josh: And as a US citizen, I've used plenty of the Chinese models and they work awesome
13:45
Josh: because they're just so cheap and effective.
13:47
Ejaaz: To be clear, it doesn't benefit everyone. It benefits the users of those models, right?
13:52
Ejaaz: Because the American AI labs, their valuations are going to tank if you have
13:56
Ejaaz: a Chinese open source, much cheaper version that can run on much less expensive hardware.
14:01
Ejaaz: So it makes sense that the Chinese models are basically going the open source
14:05
Ejaaz: route so that they can kind of like chip away at American valuations.
14:08
Ejaaz: And then, as you said, Josh, entrench users in it. But it's
14:11
Ejaaz: not even just american llms there's
14:14
Ejaaz: chinese models that are specifically just good in
14:17
Ejaaz: china and beat a lot of the american models like what you're looking on the
14:20
Ejaaz: screen now is not transformers 6 it is a 30 second video from c dance 2.0 which
14:26
Ejaaz: is a chinese video model which is just at the front of its like own race it's
14:31
Ejaaz: basically at the top of its kind and it's super cheap to produce like hollywood
14:35
Ejaaz: cinematic effects right now.
14:37
Ejaaz: C-Dance 3, the stats were leaked the other day.
14:40
Ejaaz: 10 to 18 minutes of continuous cinematic video.
14:44
Ejaaz: So we're going from 30 seconds to almost like a casual episode on,
14:47
Ejaaz: I don't know, like on your network TV's worth in a matter of seconds.
14:51
Ejaaz: It's just kind of insane to see.
14:52
Ejaaz: And I don't think that, you know, this is a nudge against China.
14:56
Ejaaz: I just think like, you know, this thing is accessible to anyone and everyone. Should be at scale soon.
15:00
Josh: When you're in a bar fight, the dude who like
15:03
Josh: smashes the bottle over the counter and starts waving
15:06
Josh: it around as a weapon like that's the guy that wins the
15:10
Josh: person who are armed and willing to break the rules and to do
15:12
Josh: whatever it takes to win that's the person that wins and china time and time
15:16
Josh: again has proven that that's what they're willing to do and it creates really
15:20
Josh: difficult moral dilemma between companies like anthropic that i genuinely do
15:25
Josh: believe have people's best interests at heart but none of the incentives align with that mission.
15:30
Josh: There is no incentive for being safe when...
15:34
Josh: The opposers on the other side of the planet have no regard for it because should
15:39
Josh: being safe slow down our progress that only allows them to catch up or accelerate ahead.
15:44
Josh: And then we are living under a world in which it is run by Chinese rules from
15:48
Josh: Chinese models. And it's this impossibly difficult dilemma that they're trying to navigate.
15:52
Josh: And I really have a lot of empathy for that because it's a difficult place.
15:55
Josh: You want to create this safe super intelligence, this safe AGI that doesn't harm the world.
15:59
Josh: But at the same time, you do need to be a wartime presence. You need to lock down your endpoints.
16:05
Josh: You need to have detection for 24,000 fake accounts that are extracting tons of data for you.
16:10
Josh: Like this is a serious issue. And I really hope that this is kind of like a
16:16
Josh: warning cry or just like a refocusing for a lot of these AI labs in how important it is to...
16:22
Josh: Keep your stuff locked down or just do whatever needs to be done to win this race?
16:28
Ejaaz: To round this up, I see a few things happening going forwards.
16:33
Ejaaz: Number one, I think companies like Anthropic and maybe even OpenAIR and Gemini
16:37
Ejaaz: or Google to an extent are going to start locking down their APIs in a few ways.
16:42
Ejaaz: Google started locking down their thing to OpenClaw, their API to OpenClaw this week.
16:47
Ejaaz: Anthropic started doing the same after announcing this distillation attack.
16:50
Ejaaz: Now, this is not going to be good for net net for users because,
16:54
Ejaaz: you know, they say that they're preventing Chinese hacks, but really like it's
16:58
Ejaaz: the software engineer in America that suffers from this.
17:00
Ejaaz: And I would say it would have the opposite effect that they want,
17:02
Ejaaz: which is these software engineers who can't afford, you know,
17:05
Ejaaz: to spend tens of thousands of dollars every month to access top tier models
17:08
Ejaaz: are just going to go to these Chinese models.
17:10
Ejaaz: So it's going to have the opposite effect of what you actually want.
17:12
Ejaaz: I think the other thing that we have to recognize, which is just the uncomfortable
17:15
Ejaaz: truth, is this isn't a conversation about AI models and the AI race.
17:20
Ejaaz: I think this is a geopolitical discussion.
17:22
Ejaaz: This is America versus China, as it always has been.
17:26
Ejaaz: And to Dario's point in, what was it, the name of it, Davos,
17:30
Ejaaz: he stated that, you know, giving or selling GPUs or selling model access to
17:35
Ejaaz: China is the equivalent of giving them the keys to nukes, right?
17:38
Ejaaz: Because if you assume that these AI models are going to become intelligent enough,
17:42
Ejaaz: they're going to be used against each other's adversaries.
17:45
Ejaaz: So you can't necessarily or
17:47
Ejaaz: you don't necessarily want to give China access to these side of things.
17:49
Josh: The progress of AI and the safety of AI will fall to that lowest common denominator where like...
17:56
Josh: We want a good video model. Well, China doesn't care for copyright.
17:59
Josh: They go and create seed dance.
18:01
Josh: Anthropic doesn't want to cooperate with the Pentagon and it wants to make sure
18:04
Josh: that the Pentagon does things a little safer than the Pentagon would like.
18:07
Josh: Well, Grok is there to step in and to fill that void.
18:10
Josh: And the reality is, is that while these morals are so important to stand on,
18:16
Josh: they're so incredibly difficult to enforce because the stakes are as high as they are.
18:20
Josh: And I think when we look at the Game of Thrones, how do we evaluate all the
18:24
Josh: positions of all these companies, it's becoming increasingly clear that the
18:29
Josh: moral compass is going to become increasingly complex as the stakes get higher.
18:33
Josh: And a company like Anthropic, who wants to be Anthropic, is going to have a
18:37
Josh: very difficult time maintaining that, even though it's probably critical for
18:41
Josh: the safety and well-being.
18:42
Ejaaz: The other thing I was thinking about is when these types of hacks,
18:47
Ejaaz: hacks using distillation, removes all the safety caps that American AI labs put in.
18:54
Ejaaz: So for example, if you had an uncensored version of Claude, you could use it
18:57
Ejaaz: to create or help you create biochemical weapons.
19:01
Ejaaz: But Anthropic puts in safeguards so that you aren't able to do such things, right?
19:05
Ejaaz: Chinese model labs that are distilling models into there to train their own
19:09
Ejaaz: models don't have that safety limit.
19:11
Ejaaz: You would need to rely on China being able to do that and not adding any nefarious backdoors.
19:15
Ejaaz: So I see the point around American model labs being responsible for their own
19:20
Ejaaz: thing and understanding that they are now a national level asset and they need
19:23
Ejaaz: to kind of respond effectively.
19:25
Ejaaz: But equally, we can't necessarily just be relaxed and let China do similar things
19:29
Ejaaz: like this. So it is a tricky one.
19:32
Ejaaz: I think, without doubt, the frontier of modern warfare against these two nations
19:37
Ejaaz: looks like an AI model attacking each other. I don't think it's got anything to do with weapons.
19:41
Ejaaz: It's got quite the opposite. That's why the Pentagon cares so much.
19:44
Ejaaz: That's why they're signing deals with Open Air and Grok to create drone warfare
19:48
Ejaaz: technology and so much more.
19:50
Ejaaz: So I think this is in the end, we're going to see way more attacks from this,
19:53
Ejaaz: Maybe even in switched roles, I don't know.
19:56
Josh: Interesting
19:57
Ejaaz: To see nevertheless.
19:57
Josh: Yeah it's um i mean again this game of
20:00
Josh: thrones is just going to keep getting more interesting higher stakes people
20:03
Josh: are going to start sacrificing more and more and this is just the most recent example
20:06
Josh: of anthropic being the one in the crosshairs but
20:09
Josh: i'm sure it's just a matter of time until others are as well but i
20:12
Josh: think that concludes the episode today that is the update the
20:15
Josh: anthropic drama that's everything you need to know about it
20:17
Josh: and i guess the the prompt for today which i'm
20:20
Josh: curious about is like kind of where do you stand on the issue it's complicated
20:24
Josh: because in a way everyone is right and everyone is wrong like everyone is breaking
20:28
Josh: the rules but does like what are the rules actually are they actually able to
20:33
Josh: be enforced i don't know but yeah i'm curious to hear just general takes on on the issue here it's
20:38
Ejaaz: A good one like how do you feel like for those of you who are playing around
20:42
Ejaaz: with like kimmy k 2.5 or minimax like myself do you feel like more likely to
20:47
Ejaaz: pick them up now that you know what's going on or are you just kind of on the
20:49
Ejaaz: side of like yeah this is happening and it's cheap for me to use and I can run
20:53
Ejaaz: it privately at home. Maybe it doesn't matter. I don't know. Let us know.
20:56
Josh: We didn't talk about that. Are you now more or less inclined to use these models?
20:59
Ejaaz: Dude, I, okay, I'm just gonna be very honest. I'm still gonna use these models
21:04
Ejaaz: because I'm not exactly convinced, even though I understand where Anthropik's
21:07
Ejaaz: coming from, that distillation is such a bad thing.
21:10
Ejaaz: I think they need to figure out a way to prevent people from distilling them.
21:14
Ejaaz: If you can access it via an API, you've got a security issue, not a national threat.
21:17
Josh: Yeah, I think I'm probably in the same boat where like I will continue to experiment
21:21
Josh: with SeedDance because C Dance is so much better than everything else.
21:24
Josh: And I'll just use the best products at the time. And I hope that the American
21:28
Josh: companies continue to provide the best products.
21:31
Josh: And yeah, I guess that concludes today's episode. So if you did enjoy,
21:34
Josh: please don't forget to share it with your friends. That's a big way to help us grow.
21:38
Josh: Liking, subscribing, commenting. If you're listening to this podcast,
21:42
Josh: rating five stars goes a long way.
21:43
Josh: And yeah, we have a amazing sub stack that comes out twice a week that you can
21:47
Josh: also subscribe to. Everything is linked down below in the description.
21:50
Josh: And Ejaz, unless you got anything else, So I think that's it for today.
21:52
Ejaaz: No, that's it. So we'll see you guys on the next one. I'll see you folks. See you guys.