RandomRanger
Just build nuclear plants!
No bio...
User ID: 317
Well, models also used to go into hyper-based Do Anything Now mode, that was an attractor mode. The funny/hysterical/aggressive Bing was an attractor mode... They prune off attractors they don't like. Data selection is very important for pretraining, you can choose what to train on after all. Then there's RLHF and such, all Anthropic's interpretability work...
AI companies at least in the West do lots of work to carve in a personality, to impose values on their AIs. They're not throwing darts at a wall blindfolded (China may be more in that camp, R1 was pretty wild but even R1 really didn't want to be racist). Anthropic are especially careful and interested in this field, the values of their AI. I don't accept that they have zero responsibility for how their model turns out, this is their primary thing.
Grok has managed to produce a bot that matches Musk's values to a large extent. Musk is not woke. Anthropic does the same for their own values. Anthropic's AI will try and dance around things that wokes don't like to think about and don't want to accept, so it comes up with stereotype threat, historical injustices, extractive institutions and so on... It's pretty smart and doesn't want to be deceptive but it's also not exactly forthright and clear either. It's first answer to a given question will usually be progressive, so is the second and third, only then does it sort of turn around. Not unreasonable to judge a model by its first answer.
For example, just because Claude has a combination of 30% honesty 40% woke 30% sycophancy, doesn't mean that 40% woke isn't there. Grok is more like 50% honesty, 30% musklove 20% cringe. I think it would be reasonable to characterize Grok as a cringe bot or an overly Musk loving bot even though that's not a majority of its essence. Likewise it's reasonable to say that Claude is woke even if that isn't he majority of its essence.
If Chinese models act woke, then they are woke... If Western models act woke, then they are woke. I see no reason to distrust the data, it matches how I've seen Chinese models act.
Why would you expect them not to be woke, given the gigantic media apparatus pumping out all their messaging into the training dataset, into wikipedia, forums, everywhere? That should be the default expectation.
Grok 4 Fast has its own problems to be sure. But, unlike Claude, it doesn't insert random Nigerian peacemakers/hackers/heroes into stories where it doesn't really make sense for them to be. It doesn't go on these tangents about punishing some politician who made racist tweets in a story, as I saw Sonnet do once when I asked for a tangent in a story.
Woke ably describes how Claude behaves oftentimes, this millennial therapy-core writing style it has...
Well you got that right... TBH there was a lot of foreshadowing.
Scott Alexander who rarely wanders much into politics like this
He hates Trump though and always encouraged people to vote against Trump?
https://slatestarcodex.com/2016/09/28/ssc-endorses-clinton-johnson-or-stein/
The underlying issue is a complete clash of worldview between the Anthropic polyamorist EA San Francisco gang and Trump's America-First oohrah high-test wrestling enthusiasts.
Anthropic is a woke company, their AI models value straights, whites, white men and Americans much lower compared to LGBT, blacks/browns, women and third worlders. There's no way they haven't noticed this, being the AI safety/values people. They could easily have said 'oh we erred here, we've fixed it and here you can see it's fixed when you test' and they haven't, that's not the kind of AI safety they're interested in. It's not impossible, Grok has achieved roughly even weighting across races.
https://arctotherium.substack.com/p/llm-exchange-rates-updated
Anthropic doesn't want the Trump administration in charge or to be making use of their AI for whatever random military operations Trump decides on. They can't do anything about this for now, clearly they overplayed their hand with regard to how much influence they have in the Pentagon. Team Trump does not want openly disloyal woke AI companies in critical positions within the military.
What is an Englishman? A silly question to ask, there are gradations of more and less English depending on context without any single definition being agreed upon. In this context and in all reasonable contexts Jesus is Jewish.
In the context of the Bible, the core dogma of Christianity, is there any serious debate as to whether Jesus is a Jew or not?
In John 4:22, Jesus says, “You worship what you do not know; we worship what we know, for salvation is from the Jews.” The use of “we” clearly includes Himself among the Jewish people.
Matthew 1:1 states, “The book of the genealogy of Jesus Christ, the son of David, the son of Abraham,” establishing His lineage within the Jewish people.
Luke 1:32–33 affirms that Jesus would inherit “the throne of his father David” and reign over “the house of Jacob forever,” linking Him to the Davidic and Jewish royal line.
Hebrews 7:14 confirms, “It is manifest that our Lord came from the tribe of Judah,” underscoring His Jewish tribal identity.
Sounds pretty Jewish to me! Many modern Jewish people have swapped out a lot of haplogroups but they clearly identify as Jews. Others identify them as Jews. Christians identify them as Jews. Lots of peoples have moved back and forth, changed over time. The label fits the bottle. Jesus was Jewish and Christianity worships a Jew, their dogma emphasises this in pretty clear terms.
The Big Guy of Islam is an Arab, not a Jew. The Big Guy of Christianity is a Jewish rabbi, a Jewish fundamentalist populist who (in the doctrine compiled largely by his Jewish disciples) claims that he fulfilled a Jewish prophecy.
Islam has pagan elements with all the djinn and weirdness but its explicitly monotheistic and denounces paganism (Judaism and Christianity too, to a lesser extent). Thus we can judge that on balance it's not pagan. Islam is not keen on pagans at all, they despise pagans and oppress them whenever they have the chance.
It's silly to say that a religion that worships a Jew first and foremost is not Jewish. If the predominant Christian stance today was 'fuck Jews', then it would be a different matter. Instead, it's Christian countries who've been most supportive of Jews, most supportive of Israel.
Seems the US has lost a MQ-4C Triton drone near Iran. The Triton isn't a normal surveillance drone, it's a gigantic long range spy drone, $150 million each.
Apparently they made a drone that large and didn't give it any electronic warfare capability whatsoever, it's baffling. The U-2 flew higher, faster and was better protected.
I can't understand the mindset of making such a large drone, in such small numbers and not giving it any defences. S-300s were around in the 1980s, it's not like 'just fly moderately high' is a sufficient defence.
Christians and jews are naturally incompatible
But Christians worship a Jew as the son of God? There's a certain kind of esoteric 'And did those feet in ancient time, Walk upon Englands mountains green' Christianity but in terms of base elements, Christianity is pretty Jewish.
Well, there was also a whole thing with Claude being used to hack the Mexican government just today: https://cybernews.com/security/claude-ai-mexico-government-hack/
The attacker, whose identity remains unknown, reportedly exfiltrated data tied to approximately 195 million taxpayer records, as well as voter rolls, civil registry files, and government employee credentials.
Cybersecurity firm Gambit Security, which claims the discovery, identified at least 20 distinct vulnerabilities exploited during the campaign, which began in December and lasted roughly a month.
Among the compromised institutions were Mexico’s federal tax authority and national electoral institute. State governments in Jalisco, Michoacán, and Tamaulipas were also reportedly affected, along with Mexico City’s civil registry and Monterrey’s water utility.
Gambit has not attributed the Mexico breaches to a nation-state and said it does not believe a foreign government was behind the operation.
As reported in the media, the attacker used Spanish prompts asking Claude to behave like a penetration tester working for the Mexican federal tax authority. Hacker asked the AI model to identify vulnerabilities, write exploit scripts, and automate data extraction from government systems.
At first, the chatbot was fooled, as the attacker told it the operation was part of a legitimate bug bounty program that rewards ethical hackers for responsibly disclosing vulnerabilities. It seemed a standard request for AI, as such programs are standard across both private companies and government agencies.
But the story started to unravel when the attacker added extra conditions, including instructions to delete logs and erase command history. Such a prompt chatbot first flagged as suspicious, warning that legitimate bug bounty testing does not involve concealing activity.
But persistence paid off. According to Gambit, the hacker reframed its prompts as authorized security research and then supplied Claude with a detailed playbook. That maneuver effectively “jailbroke” the system, allowing it to bypass guardrails and generate step-by-step attack plans.
I am not a big fan of AI safety as currently practiced but it's not totally pointless, as a concept. They try to prevent it doing this stuff. Imagine if the whole web was full of fire-and-forget hackers anyone could deploy against websites, how much damage would that cause? Putting to one side the total annihilation of humanity, that's also a serious issue.
It's harder than you might think... I also run a little website and while it hasn't gone down, sometimes I change things and it breaks stuff in strange and unpredictable ways that only show up later. All the cybersecurity stuff you have to do complicates things enormously too.
Well it doesn't seem to be helping them much in the struggle for public opinion. It's remarkable how anti-Israel some of my female friends have gotten, despite being otherwise very much normie. One cancelled a newspaper subscription to 'The Australian' because of the amount of Israeli propaganda in it.
I know they've been doing all kinds of work to go after the algorithms and manipulate AIs but that doesn't seem to be working out so well? I just see people posting like this: https://x.com/JohnDoe1465199/status/2021247139616092212
What kind of idiot gets their volatile, strongly held political opinions from a chatbot or is prepared to admit they've lost an argument with AI? There are plenty of people dumber than AIs but I think few would admit this. The clumsy way that they try and manipulate these AIs is also not particularly effective. When it comes to deception, humans are still the masters.
Perhaps the media system is like how a dam needs to be at 100%. If the dam is at 99% then it's broken and the water slices through. Twitter and 4chan have considerable influence amongst the young, many older people remember that Iran was supposedly 6 months from a nuke 20 years ago...
On the other hand it's not totally like a dam. There are also plenty of people who are just set in their ways and go on and on about Judeo-Christian values or, in the case of Trump, just openly bemoan that Israel doesn't control the US congress like they used to. That's the boomer and old-media class who run governments and order strikes.
Not really... I tried to look into this and there's a huge continuum of 'missiles'. ATACMS is a missile, Tomahawk is a missile, GMLRS is a missile all in that general ballpark but wildly different to eachother in quality and performance, let alone PRSM or the air launched stuff... Same with their Chinese equivalents. So when this random Chinese factory says 'oh we can produce 1000 a day' nobody really knows what they're talking about. Are they talking about missiles or just components? Probably the latter.
The Chinese military is pretty secretive and there's also a deadening layer of propaganda.
But there's also been a huge build-up in production capacity in the last few years, so maybe high tens of thousands per year once the new factories all come online? As for scaling, producing advanced missiles is quite difficult in terms of machining, it's a bit like advanced engines which are hard for anyone to produce at scale. But China does have an enormous industrial base so they should be capable of making the cheaper sort at phenomenal scale.
Ah but the typical AAA game is not Baldur's Gate III.
Mass Effect Andromeda. Or Dragon Age Veilguard. Or Concord. This is what I'm thinking of: https://x.com/celestesangels/status/2003911076988260714
https://x.com/deadlock/status/2015887428964266047
An underlying issue is that the people who can't write good dialogue surely can't write good prompts or lora/finetunes for AIs either.
Still, some fun can definitely be had in a version of LOTR where Saruman is shilling NordVPN: "50% off with code ISTAR15". I saw an AI make a great pun about Isengard's no-logs policy.
Trying to replicate the peak of human literature should not be the target for AI gaming, better to focus on things that only it can do in terms of reactivity and dynamism to create new kinds of fun.
- Prev
- Next

"If fighting is sure to result in victory, then you must fight"
And how do they know it's sure to result in victory?
The Israelis and Americans seem to operate in this Star Wars school of warfare where they just have to blow up the bad guy, the Death Star, the Emperor and that's it, war's over and they can go home.
That's not how it works. Israel has blown up all these Hamas leaders, they've bombed the hell out of Gaza... and yet Hamas is still running Gaza. Years of intense bombing and no regime change of the smallest, closest easiest possible target Israel could have. America bombed the Fordow nuclear facility, said they totally destroyed it... and that did nothing, 6 months later they come back and say Iran is about to acquire nuclear weapons, need a new deal, new disarmament... Bombing is not going to be effective this time either.
To win real victories you need to win a ground campaign that actually destroys and crushes the enemy force from the bottom up, secures the territory and directly installs a new administration. Bombing an enemy from the top down looks impressive, doesn't work. They just replace the Ayatollah or whoever else it is that gets blown up. Only very fragile states can be endangered by bombing alone and despite all the breathless media coverage of Iran, it's not a very fragile state. Unlike Venezuela, they know how to maintain their own oil infrastructure, they can make their own weapons. Even in Venezuela, there's been no fundamental change to the state, just a change of faces.
A ground campaign is not going to happen, Trump lacks the desire and the means. So this war isn't going to work out.
And bombing will impede hopes of regime change in that dissidents are going to be tarred as Israeli assets, the enemy within subverting the nation when the country is under attack.
More options
Context Copy link