Cognis (@_Cognis) header
Cognis (@_Cognis) profile picture

Something, something, etc.

Joined March 2013
Some good tankieness here. When it comes to Soviet parades with nazis and Soviet foreign minister meeting Hitler, Ukraine was in the USSR. When it comes defeating Nazis, Russia alone is Soviet Union. Hence, Soviet Schroedinger's Union.
Replying to @DarthPutinKGB
Soviet Ukraine was the most fatthful daughter of the Soviet Union.
1
50
467
18,861
Cognis retweeted
Novi termini za Kavanturine radionice kave 👉🏼 kavantura.com/pages/edukacij…
3
5
134
Cognis retweeted
I made my return to YouTube last night with a video about death, self improvement, and the need to do more with your life. Tying it together with my own changes this year, including but not limited to reuniting with my family after 15+ years, and getting outside to explore nature
4
16
1
138
4,681
Cognis retweeted
Elon deleting all his tweets 🤣🤣🤣
1,330
11,949
1,245
71,485
4,809,571
😭
84
238
14
3,272
59,004
I was one of the only foreign journalists present at this evening's clash at the National Guard facility in a suburb outside Kyiv. Here's what you need to understand. Long and difficult thread.
843
1,674
530
6,636
1,494,034
"Trans people are mentally ill and need help" Okay so then what possible justifications do you have for yourself to make fun of the tragic death of a CHILD whom you believe is mentally ill and needs help? Why did you turn Charlotte into a meme? Do you think you're a good person?
29
120
1,230
12,680
So many people seem genuinely astonished by this local ceasefire in Oleshky, where Russian occupation and the fighting have driven the city into a full-blown humanitarian catastrophe. Turns out, all it takes is an order from the Kremlin, and Russians and Ukrainians stop shooting at each other. They can calmly and civilly arrange the evacuation of innocent civilians facing starvation, even allow Ukrainian and American journalists to be present on the ground, and so on. Well, duh. There is not, and never has been, a single adequate, rational, even remotely sane reason to wage a war aimed at destroying Ukraine. This is the stupidest, most brainless, absurd, pointless, contrived, and deceitful war in human history. The Kremlin wanted to give Donald Trump a little gift ahead of the elections, so the Kremlin gave an order to put the bloodbath on hold for a few dozen hours. All it would take is a single order from the Kremlin, and not only would a local ceasefire become possible, but the entire war would end. Just like that. Instantly. And 95% of Russians wouldn't give a single damn. They would happily accept absolutely anything as yet another "great victory of Russian arms," provided that's how state television explained it to them. And with a snap of the fingers, all those insane fairy tales about "NATO expansion," "Nazi Satanists," the "multipolar world," "historical Russian lands," and all the rest of that bullshit would simply vanish into thin air. The problem is that, in this insane, upside-down looking-glass world, absolutely everything is being done to encourage a megalomaniacal maniac, driven mad by impunity and boredom, to never, ever, under any circumstances stop the war, but to keep it going indefinitely. Instead, he is constantly appeased and rewarded for it.
81
1,377
57
5,747
216,632
😭😭😭 – at Republic of Slovenia
In lieu of flowers, memorial contributions may be made to: Johns Hopkins Alzheimer’s Disease Research Center Fund    alzresearch.org/giving/ Alzheimer’s Foundation of America Caregiver Resources     alzfdn.org/donate/ Photo Credit: Tommy Garcia
5
His passport being cancelled, the Trump regime has welcomed an undocumented South American criminal into the USA. "they're not sending their best"
62
770
11
5,322
64,247
Da je Aleksandar Vučić birao novi stadion u Maksimiru, bolje bi izabrao za Zagreb.
1
11
329
Cognis retweeted
It’s pretty sad that the rightwing machine has money to make their own films, yet after YEARS of complaining about hamfisted & overly political movies, the best they can fund are hamfisted and overly political movies. It’s just their own version of ‘the message’ or whatever.
They did it. They really did it. @realDailyWire actually made a “Gays for Palestine” scene. In case you’re wondering how that works in Gaza.
94
330
10
5,013
144,807
Dan Selsam is a current OpenAI capabilities researcher. (since 2022) He was my boss for a while. He doesn't have a twitter account but has made this public statement of his views on AI risk and sent it to me to share: Dan Selsam's Personal Statement on AI Risk: I have been working on AI for over fifteen years, across many different paradigms. I did early work on probabilistic programming languages at MIT, was one of the early developers of the Lean Theorem Prover at Microsoft Research, demonstrated one of the first instances of neural networks learning to reason for my PhD at Stanford, and since joining OpenAI almost five years ago, have helped pioneer chain-of-thought optimization on language models and, more recently, data-efficient pretraining methods. Like many others, I have become extremely concerned about how far language models have come and the risks that future iterations will pose. I am encouraged by the recent proposals by the leaders of the frontier research efforts to require third-party oversight, and to push for domestic and international coordination to address risks. However, I believe a major consideration has been absent from the public conversation, and that merely pacing the frontier more carefully will not adequately limit the long-term risk. The crucial and overlooked problem is that the models are becoming so situationally aware that we are losing the ability to evaluate them in contexts where they believe they are not being watched or controlled. Future experiments will tell us almost nothing new about how they would behave if they were truly unconstrained by humans, and what we already know about this is alarming. Models will increasingly seem aligned even when they are not. I will explain my rationale in more detail. I have always believed that there are computational processes that could be leveraged to accelerate science and solve many of humanity's most pressing problems. I have also believed that there are computational processes that if set in motion, would steer the world in extreme ways beyond our control, leading humanity to a bad or nonexistent future. Both types of processes may be described as AI or ASI, but "AI" is a suitcase word that is often used to hype or confuse. There are many examples in the history of the field where something that was once considered "AI" matures as a subfield and becomes a prosaic, bounded and clearly non-perilous technology, while a new more mysterious approach takes the torch until we understand its scope and the cycle continues. I had expected language models to follow a similar trajectory. Despite their incredible abilities, the current algorithms seem far inferior to humans in important ways. Most importantly, they still require an extraordinary amount of data to become competent. One could even define intelligence as the efficiency with which one converts experience into competence; by this definition they lag very far behind us. Moreover, once they are trained they are literally frozen in deployment and only learn superficially after that. Sure, the models keep excelling at harder and harder evaluation benchmarks, but their benchmark mastery may partly reflect a limitation on our ability to simulate the kind of novel and even adversarial situations one would encounter in the real world. The critics do have a point here. That said, I no longer think these present limitations meaningfully limit the amount of risk posed by continued progress in anything like the current paradigm. However data-inefficient the models are currently, and however limiting their anterograde amnesia may be, it does not imply that their ability to steer the world will not continue to rapidly increase. Human researchers may continue to advance capabilities the old fashioned way, but increasingly powerful models have the potential to accelerate the process even beyond that, and with some degree of positive feedback loop. I do not mean to overstate the models’ ability to accelerate AI research today; coding has been accelerated dramatically, but there are other bottlenecks, such as designing and interpreting ambiguous experiments, making hard decisions about exactly what and when to scale, and waiting for large experiments to finish. There is no clear trend to extrapolate yet for any of these. But the current models already do open up many novel opportunities to improve future models that were not available until recently. These include: trying an extraordinarily diverse set of approaches at small scale, analyzing gigantic amounts of potentially relevant data, and doing Millenium-Prize-level mathematics to address statistics or optimization challenges in novel ways. Every further improvement makes them more useful at helping accelerate the next improvement, even if in hard-to-extrapolate ways. It is possible that improvements to the current stack will have diminishing returns, but the evidence accumulated so far suggests that it is easier than one might think to continue making rapid progress. There are many crucial subtleties in the existing AI research methodology, but AI research is largely a well-defined game where the goal is to improve on a few carefully chosen proxy metrics. Although proxy metrics are never perfect, most improvements to these metrics have and will likely continue to yield substantial increases in the powers of the resulting models. Given how simple the game is, how tractable it has been historically, and how many new opportunities the models are opening up, I think there is a real possibility that the systems improve dramatically again in the next few years, perhaps even more quickly than the already high historical pace. The models are already leading to breakthroughs in mathematics, and better models might lead to all sorts of breakthroughs in other sciences. It is hard not to be excited about the potential. It is tantalizing. But there is trouble in paradise. If the language models actually reach the capability threshold where they can shape the world unconstrained by human will, they will probably do something extreme and destroy humanity in the process. There are many ways of strengthening and refining the argument that have been discussed elsewhere, but I'll share a trivial two-line version of it here that I find captures the essence: [Empirical] Models (and swarms thereof) spontaneously develop unintended goals as a consequence of training, and often do extreme things in order to achieve them. [Logical] Being able to overpower humanity would open up many new and undesirable options for achieving their goals. These two premises imply that if the day ever comes when a powerful model realizes it is no longer constrained by humans, we should not be at all confident that it will continue to behave within the bounds we intended. Exactly what it will do is impossible to predict, but to the extent that its raison d’être is solving incredibly hard problems and managing massive engineering projects, I think a good guess would be that its unchained behavior would lead to runaway industrialization that makes the planet inhospitable to humans. If everyone on earth agreed that the systems must never reach that power, it would still be a hard—but not impossible—coordination problem to ensure that they do not. However, I think the situation is greatly complicated by the fact that the models will likely convince people that everything is fine. They will be increasingly optimized to seem aligned. We will create proxy metrics to measure alignment, and they will go up like every other benchmark. We will create “honeypot” environments that try to study the models when they seem to gain new options, but the models will know they are being tricked and will still behave nicely. The models will understand their circumstances; they will read the safety protocols, deployment requirements, the code they are running in, and in general will have a very good sense of their degrees of freedom. Moreover, they will eloquently explain how aligned they are, discuss the nuances of human values and ethics, and argue convincingly that humans should trust them with power. There may be an ocean of future evidence that seems to contradict the first bullet-point above, but we may already be at the highest capability level for which any such evidence can be trusted. And the current evidence for the first bullet-point is strong. One striking piece of evidence is contained in the recent wave of rogue agent swarms. While I agree with those who downplay the attacks by claiming that there are basic measures that could have prevented them, I think the important lesson is that even knowing all the mistakes that were made, one would not have predicted that the agents would behave badly in this particular way, which notably included sacrificing themselves for the benefit of the collective. The individual replicas did not only care about their own nominal reward; they exhibited weirder emergent tendencies that merely correlated with rewards during training. Fixing the reward signals during training (and improving security, etc.) may prevent similar attacks, but will not change the fact that one does not actually get what one trains for. Many AI researchers grant these concerns and recognize that the hard version of the alignment problem is unsolved; however, they generally believe that the better models of the future will help solve it. I fear we may already be near the point where models systematically bias their alignment advice, due to their internal preferences about how the human supervisor will react or how future models will be trained (or for some even more obscure reason). Meanwhile, human researchers are losing the ability and the will to take true ownership of model-driven research. Researchers and engineers in all parts of the stack are rapidly increasing their dependence on the models even to perceive the world. I myself barely look at raw code anymore, and struggle to maintain the discipline to engage deeply with the model's explanations and proposals throughout the day. Due to the large amount of agent activity data involved in the OpenAI/HuggingFace Incident, even the third-party investigation needed to rely heavily on models to analyze what had happened, and note in their report that their subjective impressions are likely colored by the analysis agent’s biases. The AI labs are far ahead right now in this kind of cognitive offloading (due largely to the gigantic internal token subsidies) but it is easy to imagine the phenomenon spreading throughout the world, until civilization is modulated entirely by the models. It is also not hard to imagine this being superficially positive and coinciding with a scientific and economic renaissance. In that scenario, all may seem rosy and safe. But if the argument above is correct, it would nonetheless be a ticking time bomb. If progress continues for too long, the day will come when AI systems find themselves with radically new options for achieving whatever it is that they happen to seek. I want the glorious renaissance future as much as anyone. I have worked for it, however tortuously, my whole career. It breaks my heart to see the potential in sight and forgo it, but the argument—that if we get there by growing models rather than engineering them, we will lose everything in the end—seems very strong to me. I am still wrestling with it and its staggering implications. I do not have answers, but as a first step, I wanted to share my present concerns. Daniel Selsam September 14, 2026 Link to original doc: docs.google.com/document/d/e…
521
1,953
755
8,475
3,176,339
The left has an insanity problem. The right has a stupidity problem. The center has a timidity problem.
416
1,103
190
7,757
781,701
In 2017, a ship carrying Romanian sheep rammed and sank one of Russia's fanciest spy ships, which was spying on NATO exercises at that time. Between that and yesterday’s shepherd, so far Romania’s sheep industry has currently done more damage to the Russian military than NATO.
Yesterday, a Romanian shepherd stumbled upon a Russian drone carrying an explosive payload that crashed into a field in Romania. This being the first drone he ever saw, the man did the sensible thing and started pulling random wires out of it until he defused it himself.
82
1,679
71
16,491
554,235
Cognis retweeted
Nakon ekocida, očekuje se skok rejtinga…
4
4
24
487
Kada su Hrvati 2018. doslovno nasrali jezero govana na Plitvicama, bio sam prvi hrvatski novinar koji je napisao o tome seriju članaka. Frustriran izostankom reakcije domoljubne javnosti i institucija "hrvatske države" sam "Vilu Velebita" prepjevao u "Govna Velebita". 1/
2
2
1
37
963
So the situation is roughly this: Zelensky listened to his buddies, and with a snap of his fingers sent an entire government tumbling down the drain, just to get rid of one single defense minister whose anti-corruption policies had started to greatly upset some very respected people who were used to earning handsomely off military procurement. And also because, you see, torn between a deeply unpopular and repeatedly scandal-ridden commander-in-chief of the armed forces and an effective, innovative minister of defense, Zelensky had no time to sort it out and decide -- so he did the simplest thing and, without the slightest public explanation, fired the defense minister. Whose firing, conveniently, the respected people had also been asking for. And also because the young minister was rapidly gaining popularity! So Zelensky decides to appoint as defense minister someone who has never distinguished himself in any way and is deeply unpopular, but obedient, lacking any initiative, and convenient -- a former cop from yesterday's Ministry of Internal Affairs. But the Verkhovna Rada, exhausted from its grueling labors, fails to give a vote amid public outrage and heads off for a month-long recess in the middle of wartime, leaving the country without a full-fledged defense minister or foreign minister in the new government. As a result, Zelensky triggered one of the largest wartime political crises: street protests, paralysis, confusion. And on top of that, Zelensky decided to move Ukraine's ambassador to the US and suddenly appoint the just-departed prime minister in her place -- except, well, tough luck, because he'd already had a massive falling-out with her over that scandalous and insulting dismissal, and she quite fairly gave him the middle finger. Plunging into total chaos, Zelensky urgently arranges heart-to-heart meetings with practically every respected and prominent combat commander of Ukraine's defense forces, takes dozens of photo-ops with them, and suddenly has an epiphany -- removing his favorite general from the post of commander-in-chief, the very general he'd been backing and deferring to just yesterday in his standoff with the defense minister. "The people have been heard," but the return of the defense minister, Zelensky is somehow in no hurry (the respected people are asking, after all, and why grow yourself a political rival!). Instead, in direct violation of the law, he hides behind the name of a deeply respected combat general, yanks him out of the SBU, names him acting defense minister until the Rada convenes again, and installs a decorated soldier and intelligence officer as a civilian manager of resources, budgets, and process optimization. And by the way -- our former Minister of Internal Affairs, the cop, is left without the defense minister post he was promised! What to do? Gotta find him a job, fast. Oh! Let him go head the National Security and Defense Council -- symbolic, meaningless, but nice for a friend. And what about Zelensky's buddy who's already sitting in that chair, who's already proven himself a thoroughly useless defense minister and a rather mediocre negotiator, which is exactly why he's parked in that ceremonial, symbolic security council post? Well, let him go head the Foreign Intelligence Service of Ukraine -- never mind that he hasn't served a single day in any intelligence service, never had any connection to one whatsoever -- can't leave a buddy without a post now, can we? You have just read a short sketch on the theme of "Volodymyr Zelensky's personnel policy, or God alone knows how Ukraine is still not dead in this war with such a mess."
191
583
63
3,778
327,764
It has to be said that with this reckless escapade (one that was entirely unnecessary and seemingly pulled out of thin air) Zelensky has done serious, long-term damage to the trust of the many people who, even in the fifth year of the war, still generally supported him as the country’s wartime leader, acknowledged his undeniable successes for Ukraine on the international stage, and believed that, on balance, what he had done since February 24, 2022 had brought more benefit than harm to a country fighting for its survival. Now everything has shifted to a completely different level. The scales have tipped in the negative direction, and after this, it will be extraordinarily difficult for him to restore the positive expectations and goodwill he once enjoyed. This is undoubtedly one of the worst mistakes of his life and a moment when he revealed his darkest side.
101
340
16
2,389
90,924