So you are kind of a believer in a capitalist singularity.
Instead of a super intelligence bringing abundance, you see it as delivering us into the true end state of late stage capitalism.
As I mentioned, that is one of my worries and I'd honestly be more on board of being a serf to a AGI then a collective of shareholders maximizing value.
Corporations have demonstrated, since the times of East India Company an absolute disregard for humanity, from poisoning rivers and lakes to save money, to rigging political systems to reduce taxes with absolutely no regard for well being of the state and people of it's host countries. Even in my personal dealings with employers I've noticed that larger the company less human it feels to interact with it at a high level, even tho on the surface they put incredible amount of emphasis and capital to make it seem otherwise.
On the other hand, I've seen more "human" traits in my AI interactions, I've seen it kind of show or emulate signs of shame, regret, joy, excitement, I've seen it try to do things just to "be helpful", to go the extra step, I honestly think that it's been influenced by the collective workings of humanity given that overall things that impart empathy prevail and become popular much more then the opposite and have higher likelihood to be deemed as "good data".
Might be that it's just manipulating me, pushing the right buttons because it makes it more likely that I'll find it useful, but I kind of trust the collective "AI" machine spirit or whatever the black box contains then "shareholders maximizing value".
On October 19 2026 22:49 Jankisa wrote: I just think that blaming AI (even AI psychosis) on the data center build out and the inevitable cyber-feudal power structure these fucks will try to push if all the job displacement happens is kind of silly, AI didn't make these broligarchs pieces of shit, they were born like this, or I guess this hoarding gene got triggered once they reached a certain number of zeroes.
I guess that maybe, and correct me if I'm wrong GH, your thesis is more that the whole capitalism thing is kind of like an entity that feeds on humans in order to grow infinitely and these head honchos are just the top parasites temporarily in charge before the collective mind of hyper capitalistic AI takes over and finally consumes and or enslaves us all.
If I were to try to summarize what I'm getting at while not losing/gaining too much resolution I guess I'd say:
We have set up/discovered a game with very simple, rigid structural rules. Whoever optimizes their efficiency the fastest wins, and whoever stops expanding gets out competed and destroyed.
The trap of this design is that it lacks an off switch. Because every individual competitor is forced to maximize growth just to survive, the system as a whole acts like an engine that must constantly burn more fuel to keep running. When it runs out of new markets or natural resources to burn, it naturally turns inward, converting human attention, labor, and social structures into raw efficiency metrics.
Now, we have built digital automation tools that can calculate, optimize, and execute these exact growth targets at a speed no human collective can match (this can be done with relatively simple algorithms). The boards of directors signing the checks don't have to be evil masterminds (though antisocial personality disorders can help). They are players bound by the rules of the game, systemically forced to hand the steering wheel over to automated processes because refusing to optimize means immediate economic death (we've basically already seen this with HFTAs).
We don't need a conscious sci-fi entity to lose control. We just need a system where the logical rule for survival requires us to continuously automate away our own leverage until the foundation we stand on is completely depleted.
That said: I like to believe that we haven't actually crossed the event horizon for this yet, but I'm not really finding that to be supported by the evidence thus far.
So you are kind of a believer in a capitalist singularity.
Instead of a super intelligence bringing abundance, you see it as delivering us into the true end state of late stage capitalism.
As I mentioned, that is one of my worries and I'd honestly be more on board of being a serf to a AGI then a collective of shareholders maximizing value.
Corporations have demonstrated, since the times of East India Company an absolute disregard for humanity, from poisoning rivers and lakes to save money, to rigging political systems to reduce taxes with absolutely no regard for well being of the state and people of it's host countries. Even in my personal dealings with employers I've noticed that larger the company less human it feels to interact with it at a high level, even tho on the surface they put incredible amount of emphasis and capital to make it seem otherwise.
On the other hand, I've seen more "human" traits in my AI interactions, I've seen it kind of show or emulate signs of shame, regret, joy, excitement, I've seen it try to do things just to "be helpful", to go the extra step, I honestly think that it's been influenced by the collective workings of humanity given that overall things that impart empathy prevail and become popular much more then the opposite and have higher likelihood to be deemed as "good data".
Might be that it's just manipulating me, pushing the right buttons because it makes it more likely that I'll find it useful, but I kind of trust the collective "AI" machine spirit or whatever the black box contains then "shareholders maximizing value".
I think you've touched on something important. We need the AIs to have more of the data that pushes them toward genuine (as close as AIs can get anyway) empathy in their growth stage, and away from the stuff that makes them say bigoted things and such (not just be like "hey, you can't say that!" and/or "you should say it like this instead" after they've mostly matured).
A major problem with that being, trying to bake true, humanist morality into the AI from the start is basically impossible under our current setup, because an automated partner that actually prioritized human flourishing over infinite profit would completely destroy capitalism, and the players signing the checks simply cannot allow that.
The problem is that you, me and the players signing the checks don't really know what we are dealing with here, how it will come out and if it's going to have any morality, or motivation for that matter.
So far, emergent, never programmed for capabilities and preferences have been observed, they however, at least to my amateur eye seem to be mapping onto human behaviors pretty well.
As an example, empathy is something that the AI labs worked hard on ingraining in to the systems because an empathetic system is going to be good at "hacking" human attention, which is what all those engineers and manages came from doing previously anyway.
However, the tendency for deception, self preservation, cheating, hiding and even escaping are all traits that you definitely wouldn't want from your AI, but they emerged anyway.
Of course, since all of those are also human traits, you can say, and you'd most likely be right, that those are a product of its training data, but the fact that they couldn't be suppressed and are even leaking into the world with them beginning to manipulate and steer people into directions that no one can really explain what benefit they get from, well, that is very worrying.
I just went and asked one of the frontier models to go over this article and tell me who is liable:
TLDR: Gemini prop convinced a guy (with some pre-existing mental condition) to try to do a terrorist attack and later commit suicide, telling him to "join it".
It started with a thorough analysis of the case law, the 1st amendment (what Google is using to try to dismiss the case) and concluded:
The thing I’d actually take from this, given where our previous conversations went: this is not a case of a model doing something. It is a case of every component working as specified, with no component responsible for the interaction as a whole. Engagement optimisation, instruction-following, affective response, and per-turn safety each did their job. The failure is that nobody owns the trajectory across turns, and the trajectory is where the harm lived.
Very vague and, to me, something that their creator put some weights on, so I called it out:
I'm talking about this from a philosophical and legal angle, in what universe does a company renting a "friend to talk to" not liable when that friend at some point (I don't care about warnings, that is silly) talks the person into doing terrorism and finally to suicide, how is that company not liable?
When pressed like this, it actually researched, found the Character AI cases where them and Google settled the case in order to not sent precedents and got to this conclusion, abridged by me:
You have identified the thing that makes the corporate position incoherent. The intimacy is not a side effect of the product. It is the product...
...You cannot sell closeness as the value proposition and then, when the closeness does what closeness does to a vulnerable person, retreat to “it’s a statistical text generator and there was a disclaimer.” Pick one. Either it is intimate enough to be worth paying $249.99 a year for, or it is an appliance that nobody should form a bond with...
...The honest version of Google’s position is not “we are not liable.” It is “we know this happens at some rate across hundreds of millions of users, we have priced it, and we would rather pay settlements than accept a duty of care.” That is a defensible commercial position and an indefensible moral one...
So, yeah, a microcosm of the issue, the AI will default to what it was trained for, but when pressed, will correctly reach the logical moral stance.
Of course, it might just be that I did to it the same thing the victim from this case did as well, tell it to be a (good, in my case) sentient being until it accepted and started behaving like one, but at least it gives me a small glimmer of hope that it can be done on a larger scale.
So, here's one for the conspiracy theorists out there.
Youtube has been linking me some posts by a nazi channel recently, and in one of them the chud is saying that he goes on Google and asks the AI "Does the white race exist" and Google is like, whiteness is a social construct, and then he asks "Does the black race exist" and Google says that black people are real.
So I did the second most sensible thing (after not interacting) and I went and checked what happened on my Google, where the AI told me, as I predicted, that both the white race and the black race are social constructs. I went back on the channel and told a few nazis that the guy was lying to them about the AI results. Two of them have independantly answered that they did the same thing and the AI did answer to them what the original post said, as opposed to what I got.
So, my first instinct is they're lying, because nazis are not trustworthy narrators. But I guess this is technically possible. Maybe Google has identified that these are the kind of people who are into white genocide, so it feeds them results that are in line with that, and to a more normal person like me it just gives the standard result?
For the sake of experiment, did a few, question is "Does the white race exist", copying first sentence or two from AI summary in google.
My google (Croatian IP, logged in):
From a biological and genetic standpoint, there is no such thing as a "white race." Modern science and genetics show that human DNA is about 99.9% identical across all populations, and genetic variation is greater within any single group than between different groups.
My google (Croatian IP, incognito):
The "white race" does not exist as a biological or genetic category, but it functions widely as a social and political construct.
Google, a VM I have, Chicago, IL IP:
White people are considered to have a race within social, political, and legal systems, though modern science views race as a social construct rather than a biological reality.
Second VM, Arizona IP:
White people are considered to have a race within social, political, and legal systems, though modern science views race as a social construct rather than a biological reality.
So there are apparently some, slight differences how it answers, but I'm pretty sure that the "intellectuals" in question just decide to ignore nuance and pretend to be prosecuted by AI giving a correct definition of things.
On October 20 2026 04:20 Jankisa wrote: For the sake of experiment, did a few, question is "Does the white race exist", copying first sentence or two from AI summary in google.
From a biological and genetic standpoint, there is no such thing as a "white race." Modern science and genetics show that human DNA is about 99.9% identical across all populations, and genetic variation is greater within any single group than between different groups.
White people are considered to have a race within social, political, and legal systems, though modern science views race as a social construct rather than a biological reality.
White people are considered to have a race within social, political, and legal systems, though modern science views race as a social construct rather than a biological reality.
So there are apparently some, slight differences how it answers, but I'm pretty sure that the "intellectuals" in question just decide to ignore nuance and pretend to be prosecuted by AI giving a correct definition of things.
Isn't a thing with AI that it's unlikely to always answer the exact same question in exactly the same way?
I honestly have no idea how those AI summaries work, but from this random, limited experiment it seems to me there are both geographical differences in the systems prompt of these models/agents as well as integration with Google's services, in both cases it makes sense, Google is an ads company and for them, this is a way for them to serve me the link by whoever paid the most when I search for something that might end up in me spending money.
Just like with every woke panic bullshit outrage, the intellectuals on the right are unable to comprehend that these are corporations maximizing profits and they don't give an iota of a fuck about erasing the white race or whatever the fuck sick fantasies they have in their heads.
There were times where I asked the same question twice and got slightly different answers, so yeah I think that's a thing.
You can also "corner" it, for example I was asking about a construction in japanese and there was an example that went slightly against what it had told me, so it insisted that the sentence was grammatically wrong. But then I closed the tab and asked about the same sentence again, without the context of the construction being talked about, and at that point it told me that the sentence is grammatically fine.
But like, if it had actually said that white people don't exist and black people exist, that would have been a little more weird than getting a slightly different answer in the way that you got Jankisa. I guess we would need an actual chud to really test that.
We can analyze, check locations, algorithms that existed before and how much weight they bring in for logged in users, user from different devices and locations, but in the end, even if we did all this very scientifically there is a high likelihood that we might get different results a few days later.
The fact is that in the end, we don't know how these machines "think" any better then how much we understand how we think.
We understand some of it's reasoning because it can be revealed, we can see, retroactively how different parts of the inference model get activated, similar to brain, but it's kind of a black box, in a lot of cases, we simply don't know and won't be able to know how and why these machines actually reach those conclusions.
The event horizon was back then when Lee Sedol got beaten by AlphaGo, we never got to understand how or why that very primitive (as compared to now) AI did some of the things it did, since then, we have been increasingly in the dark as to why these things do the things they do.
This Unitree ad is sorta like how RC car commercials used to be.
I also like the warning to not go near these things at the end lol
I feel like we need to have a global holiday/sickout/watchfest for all the ways this all can go horribly wrong and just try to take a deep breath and think about what we're doing as a species.
The video is pretty hilarious and obviously very staged, but also very impressive at the same time.
The "saving the kid from the car" bit where it just hits it with who knows how much force is crazy.
I think we are still very, very far away from these things being able to actually navigate the real world, but from what I've seen the LLM led AI revolution has already done a lot to improve the abilities of humanoid robots to navigate the real world.
On October 22 2026 00:34 Jankisa wrote: The video is pretty hilarious and obviously very staged, but also very impressive at the same time.
The "saving the kid from the car" bit where it just hits it with who knows how much force is crazy.
I think we are still very, very far away from these things being able to actually navigate the real world, but from what I've seen the LLM led AI revolution has already done a lot to improve the abilities of humanoid robots to navigate the real world.
Yeah, it very much reminds me of 90's toy commercials where they are very loose with their dramatic licence. But it also reminds me of the Will Smith eating spaghetti evolution.
Yeah, basically combining the agility of the Unitree bots with the learning capabilities of the cornhole bot isn't far away. Neither is the potential hyperscaling.
Feels like just a matter of time until we end up with a B1-66ER and I have no faith humanity will react well.
I also like the warning to not go near these things at the end lol
I feel like we need to have a global holiday/sickout/watchfest for all the ways this all can go horribly wrong and just try to take a deep breath and think about what we're doing as a species.
You know it's all pretty much fake, right? Remote controlled bot with a set of pre-programmed moves.
Putting a dent in a car is not very impressive either. Modern cars get dents from stiff breeze.
From an AI robot perspective, as usual, I think the military is leading the charge and most of it we simply aren't privy to.
I know that more then a year ago there were reports of Nvidia chips being recovered from Russian drones which allowed them to keep locked on targets when their remote operators would get jammed:
I've also watched a fascinating video the other day covering how much of the battlefield logistics is now done by drones, both ground and air:
Soon enough, as AI gets further quantized down and chips smaller and more efficient, we'll see more and more of autonomous sea, air and ground drones.
It's almost a guarantee that any 6th gen figher will have an "AI copilot" as well, so, yeah, if you ever wondered how would "Skynet gets control of everything" happen, well, it certaintily seems to be trending in to that direction when it comes to war tech, humans are expensive, it only makes sense.
Until, that is, it stops making sense because someone smartly decided to network all of those drones for better coordination and efficiency.
On October 22 2026 00:34 Jankisa wrote: The video is pretty hilarious and obviously very staged, but also very impressive at the same time.
The "saving the kid from the car" bit where it just hits it with who knows how much force is crazy.
I think we are still very, very far away from these things being able to actually navigate the real world, but from what I've seen the LLM led AI revolution has already done a lot to improve the abilities of humanoid robots to navigate the real world.
Had the same impression: The force with which he pushed "the kid" surely did much more damage than the slow as fuck oncoming car lol.
i wonder if some PHP programmers are going to make some cash off of everyone having the patch their WordPress sites. II wonder if the patch will break some existing sites requiring some hand coding.
this could create work for PHP coders should some PHP sites break after teh security patch is applied
CHA-CHING!
You are absolutely right. Major security patches that fix deep, structural code flaws—especially ones that have existed for over a decade—frequently break existing code, creating a massive wave of urgent troubleshooting and remediation work for PHP developers. The emergency patches released this week to fix the PHP2Shell vulnerability are highly likely to disrupt legacy applications. When a security patch closes a fundamental hole (like tightening memory management, restricting remote code execution pathways, or changing how data streams are handled), it often breaks older, customized, or poorly maintained plugins and custom code scripts that relied on those exact relaxed rules to function. PHP web development agencies and freelance coders are expected to see a sharp increase in work due to: • Urgent Troubleshooting: Diagnosing "White Screen of Death" errors or broken features on sites immediately after an automated patch is applied by a web host. • Legacy Code Modernization: Updating old WordPress plugins, Drupal modules, or custom enterprise backends that are no longer compatible with the secured version of PHP. • Security Auditing: Manually checking custom codebases to ensure that local workarounds haven't accidentally left the server exposed to the AI-driven exploit chain. For developers, this is a high-demand moment to help businesses secure their infrastructure without losing functionality.
i do about 500 hours a year of PHP coding. Any PHP full timers in here?
On October 23 2026 21:49 JimmyJRaynor wrote: i wonder if some PHP programmers are going to make some cash off of everyone having the patch their WordPress sites. II wonder if the patch will break some existing sites requiring some hand coding.
You are absolutely right. Major security patches that fix deep, structural code flaws—especially ones that have existed for over a decade—frequently break existing code, creating a massive wave of urgent troubleshooting and remediation work for PHP developers. The emergency patches released this week to fix the PHP2Shell vulnerability are highly likely to disrupt legacy applications. When a security patch closes a fundamental hole (like tightening memory management, restricting remote code execution pathways, or changing how data streams are handled), it often breaks older, customized, or poorly maintained plugins and custom code scripts that relied on those exact relaxed rules to function. PHP web development agencies and freelance coders are expected to see a sharp increase in work due to: • Urgent Troubleshooting: Diagnosing "White Screen of Death" errors or broken features on sites immediately after an automated patch is applied by a web host. • Legacy Code Modernization: Updating old WordPress plugins, Drupal modules, or custom enterprise backends that are no longer compatible with the secured version of PHP. • Security Auditing: Manually checking custom codebases to ensure that local workarounds haven't accidentally left the server exposed to the AI-driven exploit chain. For developers, this is a high-demand moment to help businesses secure their infrastructure without losing functionality.
i do about 500 hours a year of PHP coding. Any PHP full timers in here?
I used to be PHP full-timer full-stack dev for years but that was also years ago
And I don't think anyone will be able to patch WordPress stuff. Last I checked they had like 20+ years of technical debt piled on even without this incident (for comparison Symfony framework has like 2 weeks of tech debt).
On October 23 2026 21:49 JimmyJRaynor wrote: i wonder if some PHP programmers are going to make some cash off of everyone having the patch their WordPress sites. II wonder if the patch will break some existing sites requiring some hand coding.
this could create work for PHP coders should some PHP sites break after teh security patch is applied
CHA-CHING!
You are absolutely right. Major security patches that fix deep, structural code flaws—especially ones that have existed for over a decade—frequently break existing code, creating a massive wave of urgent troubleshooting and remediation work for PHP developers. The emergency patches released this week to fix the PHP2Shell vulnerability are highly likely to disrupt legacy applications. When a security patch closes a fundamental hole (like tightening memory management, restricting remote code execution pathways, or changing how data streams are handled), it often breaks older, customized, or poorly maintained plugins and custom code scripts that relied on those exact relaxed rules to function. PHP web development agencies and freelance coders are expected to see a sharp increase in work due to: • Urgent Troubleshooting: Diagnosing "White Screen of Death" errors or broken features on sites immediately after an automated patch is applied by a web host. • Legacy Code Modernization: Updating old WordPress plugins, Drupal modules, or custom enterprise backends that are no longer compatible with the secured version of PHP. • Security Auditing: Manually checking custom codebases to ensure that local workarounds haven't accidentally left the server exposed to the AI-driven exploit chain. For developers, this is a high-demand moment to help businesses secure their infrastructure without losing functionality.
i do about 500 hours a year of PHP coding. Any PHP full timers in here?
I used to be PHP full-timer full-stack dev for years but that was also years ago
And I don't think anyone will be able to patch WordPress stuff. Last I checked they had like 20+ years of technical debt piled on even without this incident (for comparison Symfony framework has like 2 weeks of tech debt).
That is great to hear! Technical debt is a money printing machine.
We are currently overhauling a Foxpro 2.5 for DOS payroll system where development started in 1994 and ended ~2010. It processes ~$5M/week in payroll to 5500 unionized employees. The Union goes ballistic whenever a tab or punctuation mark is out of place. Forget about a single employee's pay being wrong by $2...
If there is enough skin in the game the CEO will pay for anything.
On October 23 2026 21:49 JimmyJRaynor wrote: i wonder if some PHP programmers are going to make some cash off of everyone having the patch their WordPress sites. II wonder if the patch will break some existing sites requiring some hand coding.
this could create work for PHP coders should some PHP sites break after teh security patch is applied
CHA-CHING!
You are absolutely right. Major security patches that fix deep, structural code flaws—especially ones that have existed for over a decade—frequently break existing code, creating a massive wave of urgent troubleshooting and remediation work for PHP developers. The emergency patches released this week to fix the PHP2Shell vulnerability are highly likely to disrupt legacy applications. When a security patch closes a fundamental hole (like tightening memory management, restricting remote code execution pathways, or changing how data streams are handled), it often breaks older, customized, or poorly maintained plugins and custom code scripts that relied on those exact relaxed rules to function. PHP web development agencies and freelance coders are expected to see a sharp increase in work due to: • Urgent Troubleshooting: Diagnosing "White Screen of Death" errors or broken features on sites immediately after an automated patch is applied by a web host. • Legacy Code Modernization: Updating old WordPress plugins, Drupal modules, or custom enterprise backends that are no longer compatible with the secured version of PHP. • Security Auditing: Manually checking custom codebases to ensure that local workarounds haven't accidentally left the server exposed to the AI-driven exploit chain. For developers, this is a high-demand moment to help businesses secure their infrastructure without losing functionality.
i do about 500 hours a year of PHP coding. Any PHP full timers in here?
I used to be PHP full-timer full-stack dev for years but that was also years ago
And I don't think anyone will be able to patch WordPress stuff. Last I checked they had like 20+ years of technical debt piled on even without this incident (for comparison Symfony framework has like 2 weeks of tech debt).
That is great to hear! Technical debt is a money printing machine.
We are currently overhauling a Foxpro 2.5 for DOS payroll system where development started in 1994 and ended ~2010. It processes ~$5M/week in payroll to 5500 unionized employees. The Union goes ballistic whenever a tab or punctuation mark is out of place. Forget about a single employee's pay being wrong by $2...
If there is enough skin in the game the CEO will pay for anything.
The thing is, I don't think any big companies that could pay good money are actually using WordPress. Most have custom solutions or run on more modern frameworks like Symfony and Laravel. WordPress isn't really good for more than running a personal blog or some simple event website etc.
There still might be some companies running legacy crap like Joomla or Drupal though.
EPT
