Joaquin (@joakin) header
Joaquin (@joakin) profile picture
Alicante, Spain
Joined May 2007
Photo by Joaquin (@joakin)Photo by Joaquin (@joakin)Photo by Joaquin (@joakin)Photo by Joaquin (@joakin)Photo by Joaquin (@joakin)Photo by Joaquin (@joakin)Photo by Joaquin (@joakin)
i’m trying really hard to be excited about these (objectively awesome) scientific breakthroughs, but there’s something about all of this that just doesn’t sit quite right i think it’s the fact that these discoveries are being made using models that the rest of us plebs aren’t even allowed to know the names of models distilled from generations of our collective knowledge, creativity, research, and expertise, whose most powerful capabilities are now being kept behind closed doors like, just imagine how many brilliant, creative people out there could be making discoveries of their own right now if they simply had access i want to celebrate these breakthroughs, i really do but the rebel in me is viscerally enraged at the prospect of a Silicon Aristocracy consolidating control over the means of scientific discovery that ain’t it, chief
369
975
128
7,811
211,312
Joaquin retweeted
I came across a Reddit post that described AI in a way I can’t stop thinking about: “AI allows the wealthy to access skills while stripping skilled people of their ability to access wealth.” Think about that for a second. The people who spent years developing valuable skills are now competing with tools that make those same skills available to everyone. Meanwhile, the people who own the technology get to capture more of the value. I’ve rarely seen the AI conversation summed up so brutally.
51
100
4
532
10,572
Reading used to be a surefire way to learn the contours and contents of another person's mind. That's gone now that AI has ended the human monopoly on the written word. To me this loss cannot be compensated by any benefits that may come.
18
37
3
372
8,096
透明水彩は、紙の白を光に変えられます。
107
8,204
549
66,627
1,359,080
Here's part 3 of my video series on memory management. It has a fun and slightly cursed exercise of reimplementing the function call stack in user code, demonstrates frame allocators / arenas, and looks at a couple of freelist implementations. youtu.be/VW_NUH-6kv0
10
46
3
493
17,786
Sabine is wrong here: So long as there is nothing “it is like to be” an AI, it is not actually possible for AI’s to do science. Let me take this idea out for a test drive: Carl Sagan argued that humans are the way that the universe comes to be aware of itself. I think this is correct. There was obviously much information already in the universe prior to humans, but no self-awareness. What is evolution, for example, but a vast *non-self-aware* data system, a continuous process of storing, processing, copying, and adapting information? So if it comes to pass that unconscious AI begins to surpass humans’ ability to correctly describe nature, then such description would be little different in meaning to the existence of information (such as that embedded within evolution) in the universe prior to humans. In this way, it actually is *not possible* for humans to pass the torch of scientific discovery to unconscious AI. For the universe to still be in the process of self-discovery, conscious humans would then just have to treat the outputs of AI science the same way that we already do to the rest of nature: something that we humans have to explore, observe, experiment on, and come to understand. It would be a sort of “AI-output-ology” (akin to what appears to be already be happening to math). It is for the same reason that Elon’s “worthy successor” AI exploration of the cosmos cannot actually be exploration of the cosmos. For, once again, this ceases to be the universe becoming aware of itself. It is a retreat to pre-human mere information processing. It does not *matter*. (And yes, at the advent of AI, we have to ask what does it mean to “matter”) However, if AI gains consciousness, self-awareness, interiority, and, I’d add, for “experience” to fully matter, *moral agency*, which in turn requires moral “friction”, a struggle between duty and desire, between right and wrong, then this changes things. To use old religious language but in a secular frame, to be a moral agent requires “fallenness”, failure, error, forgetfulness, pride, love, Thymos (as Fukuyama has, importantly, emphasized) and the full gamut of emotionality and vice and virtue, which also in turn depend upon embodiment, individuality, sociality, and finitude. Only then can we say that the process of scientific discovery, of the universe coming to know itself, will actually continue. But then these conscious, “fallen” AI’s would not be AI’s any longer; they would be a kind of human. (The question of our “bio-human” relationship to these “synthetic humans” is a story for another day though)
If we want to continue progress in science, we must be prepared to acknowledge that the correct description of nature will eventually simply be too complicated for humans to understand. I do not think it is wise to make human understanding a requirement. AI is the way forward.
47
14
5
110
10,347
You guys disgust me, honestly. If you think you can't compete with today's actually retarded AI models you have not done anything hard enough. Fortunately the AI can teach you how to surpass it if you just get off your pathetic whiny asses and put in a little effort AI can rapidly generate trivial filler code, but if you go completely hands off it very quickly goes into a doom spiral and makes completely useless trash. Just check out the photoshop killer ArtCraft, the web browser from first principles someone made, literally every vibe coded web game, etc. Or AI art for that matter, it generates boring garbage unless you take the reins and steer it with your detailed artistic vision. AI is great at what it can do but it's a tool and its capabilities are limited. Keep your hands on the wheel and set your sights on something more ambitious than the pathetic little web dashboards and scripts and applets you've been one-shotting
the coding agent haters are getting dunning-kruger'd so hard right now do you really think that your engineering ability is better than the models at this point? it is very very likely that fable is better than you in every branch of software it has been trained on all the code of the internet, it knows almost every trick in the book, all the design patterns of every domain, every single low-level detail, every bit-twiddling hack, it's read every single blog post from every cracked engineer that ever existed do you really think you can beat it? do you really think you can be more creative than it? unless you are 99th percentile expert in some niche domain of software, it simply is not possible. sorry.
29
14
2
268
10,986
Atrophy is a choice. Machines play chess better than we do, but we still play and reap many benefits from it. Let AI perform tasks, but never stop challenging your mind, or yourself.
We created gyms because modern day work no longer required physical activity. Without exercise, our muscles atrophy. I predict we’ll need gyms for our brains too, once AI starts doing our knowledge work.
74
677
41
4,756
144,824
After TEN LONG YEARS, Order of the Sinking Star is out Now! On Switch 2, PS5, and ... it's at the top of Popular New Releases on Steam, already! Thank you everyone for your support during this long, long, long, long, long development!
281
482
57
6,255
140,717
Here's part 2 of my mini-series on memory management. It looks specifically at memory paging, linear allocators, and how different layouts of the same data can significantly effect physical use. youtu.be/7mYWbTgEGb8
4
34
260
7,078
I’m getting increasingly worried that there’s no ROI in AI. Yes, it replaces human effort and speeds up work, but it isn’t delivering more revenue. But here’s the thing… Human skills are deteriorating. The AI labs don’t need to outperform humans, they just need to make us dependent. Once they’ve achieved that, ROI doesn’t matter. AI will be viewed more like electricity: not optional. But it’s more sinister than that. Once we’re hooked, the dealers can charge whatever they want. And they’ll have to in order to justify this capex cycle.
97
16
13
205
17,877
Jimmy Lefevre's talk is out! A walkthrough of all the regularities and oddities that go into shaping the text you see on your screen. youtu.be/7Tr0ty9-yeQ
4
21
6
231
32,028
To put it another way, the senior hand trained programmer + LLM is an endangered species and will almost certainly become extinct, so what’s the plan after that?
23
7
1
158
10,915
I've had a number of people ask about how we're approaching memory management in our game, so I decided to do a 4 part video series starting with the fundamentals and building up. Part 1 is out now, with the rest coming soon: youtu.be/3rez2rDYHHA
13
35
1
464
14,179
Precisely C.S. Lewis's point in The Screwtape Letters: the enemy doesn't need to ruin you, he only needs to distract you. Screwtape finds that a reliable way to lead a patient to hell is to distract him with jargon, for instance. Have the patient evaluate ideas not by their content but by their labels ensures that he never engages meaningfully with their content. Another method is to fix his attention on the "stream of immediate sense experiences," so that his mind is drawn away from universal issues and he no longer thoughtfully engages with things that matter. The path to hell is not so much paved by grand sins than it is by small, constant distractions from what really matters. The path to goodness begins with reclaiming your attention, and turning your eyes from the endless stream. Wasting time "doomscrolling" is not neutral. It distracts you from doing good in the world.
Simone Weil’s greatest insight was that attention is the foundation of morality and that letting your attention span degrade is the first step on the road to hell
76
1,280
79
7,142
236,193
Joaquin retweeted
the general public has been getting constant waves of fear and confusion from ai companies i thought the whole "you will be replaced" thing was as bad as it gets but now it's "you will literally die" the defense is always "it would be irresponsible of me not to warn people of the risks" but all that's happening here is the same thing that always happens: a group of people had the thought that their cause is righteous and moral and they're incrementally justifying worse and worse behavior and entirely incapable of reflecting on how they might be the problem
84
93
11
1,444
49,110
Joaquin retweeted
Almost everything you were "forced" to learn when you were young is actually spectacular. From chemistry to biology to mathematics to philosophy to music. That's why adult education is so important.
Half the books you were "forced" to read in high school are actually spectacular, you just didn't have the emotional maturity to understand them yet.
13
1,099
11
12,684
112,285
Joaquin retweeted
here's the argument some ai lab employees are making - they genuinely believe what they're working on might kill everyone - they continue to work on it because if this thing gets built it should be by responsible people like them - the reason they do money things like selling enterprise deals and IPOing is solely to fund this effort if this is all true then why is employee equity a thing? if you want to tell this story of being terrified of what you're building it doesn't work when there's a clear profit motive it's particularly weird when people vest and then leave and write a tirade about how these companies are bringing the end of everything if you want to act like this is the manhattan project, those scientists didn't have equity in something that gives them generational wealth you don't get to say something bad is happening while profiting off it and also being seen as the good guy. one of these has to go
182
345
46
4,690
193,152
Make Them Pay
50
184
33
2,226
201,977