Rafael

6.9K posts

Rafael

Rafael

@sohakes

This was supposed to be a professional twitter but I'm so bad at this, I'm sorry, don't follow me for tech. Dev, studied some RL. Gay, he/his. Tweets in pt/en.

São Paulo, Brazil Katılım Mayıs 2018
1.9K Takip Edilen78 Takipçiler
Rafael
Rafael@sohakes·
@DioLarmguy @YOjawnZiLLA Yeah, I think it is important info, but you wrote like it was a counter argument, while I think it just adds another aspect to the same idea.
English
1
0
0
33
Dio Larmguy
Dio Larmguy@DioLarmguy·
@sohakes @YOjawnZiLLA I think it's leaving out a pretty big part of the picture to not mention it, and it's not a phenomenon constrained to this one game. You see it with other countries sometimes, but it's extremely common with China.
English
1
0
0
50
Rafael
Rafael@sohakes·
@YOjawnZiLLA @DioLarmguy Yeah, it was the quoted tweet point as well and I agree. Was just saying that "reviews are negative because of chinese reviews" fit the explanation as well, so it is not a counter argument, it is the same phenomenom.
English
1
0
0
104
Rafael
Rafael@sohakes·
@ContraPoints I think I have too much myopia for this low yellow light bullshit. Make my house an office, I want to pour light in everything.
English
0
0
0
407
Rafael
Rafael@sohakes·
@TheZvi Thought the pro tier was kinda bad with the weekly usage though. It seems the free one has no limit? Anyway I guess that is a claude and not an opus 4.7 problem.
English
0
0
0
49
Rafael
Rafael@sohakes·
@TheZvi Subbed because gemini is getting too stale. For now mostly used for day to day things and not code, and it is way better than gemini for the same prompts (some exercise, nutrition and medical stuff, some financial stuff, some hobbyist research ones).
English
1
0
0
110
Zvi Mowshowitz
Zvi Mowshowitz@TheZvi·
Claude Opus 4.7 reaction thread, it's that time again.
English
70
1
117
43.1K
Rafael
Rafael@sohakes·
@indianjonesy @ContraPoints What the fuck are the leftists here having oh dear stop using crack and hitting your head please it is bad for your brain
English
0
0
2
118
Natalie Wynn
Natalie Wynn@ContraPoints·
Getting accused of genocide denial every day because of a year-old post where I say that genocide is happening has been like cancellation exposure therapy. I can just say whatever now, I’m free.
Seph@Seph_librul

@ContraPoints @ContraPoints At least now when you make based post after based post you don't delete them right after.

English
131
150
5.6K
177.9K
Rafael
Rafael@sohakes·
@hutchinson For those who are unaware about the post, had to find it, and actually it is very good. reddit.com/r/ContraPoints… Makes it clear most people in these replies are heavily misrepresenting her or just lack basic text interpretation skills.
English
0
0
3
378
Hutch
Hutch@hutchinson·
"Is Israel committing a genocide in Gaza? Yes. Do I oppose it? Yes. Do I feel angry about it? Yes." She said that to open her I/P post. But she went on to dissect all the ways she felt the online left sabotaged their own activism, so I guess this means she's "pro-genocide."
Hutch tweet media
English
130
231
4.7K
381K
Rafael
Rafael@sohakes·
@marshylul @hutchinson There were no physical attacks on jewish people specifically AFAIK, but there were fights with people from Israel, and the majority of people feels it's deserved because "they are all war criminals". I... I don't think this help Palestine in any way.
English
0
0
1
50
Rafael
Rafael@sohakes·
@marshylul @hutchinson I mean... that's exactly what is happening, I feel, and it kinda plays in Israel playbook. I see it in real life and in online communities in my country, leftists are turning into anti-semites and not realizing how problematic it is, because the hate feels deserved.
English
1
0
9
527
Rafael
Rafael@sohakes·
@vkrajacic I think for actual code issues you can use to test hypothesis, it can check the code way faster than you if you are looking into a very big system. I think it was not trained to debug so it is not great at it though. Same on a/b tests, it is just not very good with statistics.
English
0
0
0
46
Vjekoslav Krajačić
Vjekoslav Krajačić@vkrajacic·
So I just debugged a problem that took me a while to figure out. It required inspecting the code through various phases in the debugger. And it made me wonder: what does a coder who relies heavily on AI tools do in cases like that? After AI has given them wrong results after countless "fix this" queries? At what point do you accept that AI won't solve it for you? What if you don't know or don't care what a debugger is? Honest questions btw.
English
103
1
233
27.6K
Rafael
Rafael@sohakes·
@ContraPoints What to do when the world proves an egomaniacal narcissist right though
English
0
0
0
94
Natalie Wynn
Natalie Wynn@ContraPoints·
His entire career has been about emblazoning his name and image on vast architectural erections, golden idols, pharaonic monuments, religious icons, US currency, the Moon, history itself. You poor drooling imbeciles have been worshipping a cosmological narcissist.
Natalie Wynn tweet mediaNatalie Wynn tweet media
English
43
241
3.5K
68.1K
Rafael
Rafael@sohakes·
@KarimBaziz4 @esjesjesj Damn AI is not great for the minds of people. You are not a bluecheck to be baiting, so you are just unable to read :(. Wish you get well.
English
0
0
0
254
evan loves worf
evan loves worf@esjesjesj·
Because the right wing dictator was killing socialists
evan loves worf tweet media
English
159
2.4K
45.6K
417.5K
Rafael
Rafael@sohakes·
@Authw8 @ManTicktoc41268 I'm unsure actually. A lot of people I know who liked the game are not even leftists, most are just "apolitical normies". Anedotical evidence of course, but the game was quite popular in gaming circles, and gamers mostly share Andy Weir opinion on media.
English
0
0
0
14
tom bombadil
tom bombadil@Authw8·
@sohakes @ManTicktoc41268 the problem with disco elysium is it's preaching to the choir. the people who like it were already inclined to agree with the creators on most topics. high impact cultural change comes from persuading people who aren't already inclined to agree with you
English
2
0
0
43
tom bombadil
tom bombadil@Authw8·
it's annoying to me that people keep criticizing andy weir's approach to politics and social commentary without even noticing that this attitude has led to a massive propaganda victory for the ideas andy weir specifically thinks are cool and interesting. storytellers don't usually change the world by setting out to change the world. they change the world by focusing on their craft and letting the story speak for itself. it's basically impossible to write a whole book without injecting your views into it, but the chiller you are about it, the better it works.
Fandom Pulse@fandompulse

Project Hail Mary writer Andy Weir on social commentary in books: "I dislike social commentary. Like… I really hate it. When I’m reading a book, I just want to be entertained, not preached at by the author. Plus, it ruins the wonder of the story if I know the author has a political or social axe to grind. I no longer speculate about all possible outcomes of the story because I know for a fact that the universe of that book will conspire to ensure that the author’s political agenda is validated. I hate that." "I put no politics or social commentary into my stories at all. Anyone who thinks they see something like that is reading it in on their own. I have no point to make, and I’m not trying to affect the reader’s opinion on anything. My sole job is to entertain, and I stick to that." "To that end, I also don’t talk about my personal political opinions publicly. I don’t want readers to even know, honestly. I don’t want that in the back of their minds as they read my stuff." Is this why he has the #1 sci-fi movie in decades?

English
60
132
2.4K
181.8K
Rafael
Rafael@sohakes·
@Authw8 @ManTicktoc41268 Same for disco elysium, people who love it really do love it. People learn and get into politics because of it. Meanwhile people like project hail mary the way they like marvel movies. Nothing bad about that, it is a cool movie, but you can't deep drink from it.
English
1
0
1
72
Rafael
Rafael@sohakes·
@Authw8 @ManTicktoc41268 Wow, that is a very strong statement! Hail mary influence will be like, a little more people into physics maybe, while even if fewer people read infinite jest, the influence is way more farther reaching simply because the people who read it will have a way stronger impression.
English
3
0
1
90
Rafael
Rafael@sohakes·
@purplepain_ @DukeOfZamunda Well the bar is usually negative, I think I would love if some politicians in my country just did nothing and collected their paychecks. Just walk above the earth and you are fine, but a lot of them are little imps that fuck up public goods for their personal benefit.
English
0
0
1
182
NFL YOUNGBOY
NFL YOUNGBOY@purplepain_·
@DukeOfZamunda Fixed some potholes and walked a few miles and he’s everyone’s favorite politician all of a sudden bar can’t be that low
English
30
0
74
20.4K
Rafael
Rafael@sohakes·
@PanasonicDX4500 Extremely normal tweets for a pope now called woke because the overton window now is almost centered in "genocide is ok" on the right. Quite sad situation the world is in now, I don't get it.
English
0
0
3
265
Rafael
Rafael@sohakes·
@emollick You are the person who probably has the most evidence here, but are you sure? It does feel like the verifiable domains are advancing way more rapidly than the rest, even though the rest is still improving.
English
0
0
0
20
Ethan Mollick
Ethan Mollick@emollick·
And all the evidence is is that is that models are getting better all this other stuff at the same time as they are improving in coding. More recent models are more creative, for example. Still plenty of jaggedness, but the frontier moves more in synch than we might expect
English
3
1
56
8.6K
Ethan Mollick
Ethan Mollick@emollick·
AI is jagged, but I think sometimes it is easy to overly focus on that. The generalness is a surprise too! LLMs may be optimized for verifiable fields like coding, but AI is also not bad at corporate strategy & medical advice & writing a sestina & expressing empathy & ideation.
Andrej Karpathy@karpathy

Judging by my tl there is a growing gap in understanding of AI capability. The first issue I think is around recency and tier of use. I think a lot of people tried the free tier of ChatGPT somewhere last year and allowed it to inform their views on AI a little too much. This is a group of reactions laughing at various quirks of the models, hallucinations, etc. Yes I also saw the viral videos of OpenAI's Advanced Voice mode fumbling simple queries like "should I drive or walk to the carwash". The thing is that these free and old/deprecated models don't reflect the capability in the latest round of state of the art agentic models of this year, especially OpenAI Codex and Claude Code. But that brings me to the second issue. Even if people paid $200/month to use the state of the art models, a lot of the capabilities are relatively "peaky" in highly technical areas. Typical queries around search, writing, advice, etc. are *not* the domain that has made the most noticeable and dramatic strides in capability. Partly, this is due to the technical details of reinforcement learning and its use of verifiable rewards. But partly, it's also because these use cases are not sufficiently prioritized by the companies in their hillclimbing because they don't lead to as much $$$ value. The goldmines are elsewhere, and the focus comes along. So that brings me to the second group of people, who *both* 1) pay for and use the state of the art frontier agentic models (OpenAI Codex / Claude Code) and 2) do so professionally in technical domains like programming, math and research. This group of people is subject to the highest amount of "AI Psychosis" because the recent improvements in these domains as of this year have been nothing short of staggering. When you hand a computer terminal to one of these models, you can now watch them melt programming problems that you'd normally expect to take days/weeks of work. It's this second group of people that assigns a much greater gravity to the capabilities, their slope, and various cyber-related repercussions. TLDR the people in these two groups are speaking past each other. It really is simultaneously the case that OpenAI's free and I think slightly orphaned (?) "Advanced Voice Mode" will fumble the dumbest questions in your Instagram's reels and *at the same time*, OpenAI's highest-tier and paid Codex model will go off for 1 hour to coherently restructure an entire code base, or find and exploit vulnerabilities in computer systems. This part really works and has made dramatic strides because 2 properties: 1) these domains offer explicit reward functions that are verifiable meaning they are easily amenable to reinforcement learning training (e.g. unit tests passed yes or no, in contrast to writing, which is much harder to explicitly judge), but also 2) they are a lot more valuable in b2b settings, meaning that the biggest fraction of the team is focused on improving them. So here we are.

English
22
19
392
40.8K
Rafael
Rafael@sohakes·
@StatisticUrban Libertarians don't get it, I tried. "Look at all the horrible shit that will come to pass if the world functioned like you do". They first argue that the world would be fine, but failing that (it always fail) they just turn back to their flawed moral axioms.
English
0
0
0
146
Hunter📈🌈📊
Hunter📈🌈📊@StatisticUrban·
At a certain point you simply have to say to the libertarians "no, wrong." An orphaned baby does in fact have the human right not to starve to death in the street. Any form of society that can't guarantee that right is bad.
edelnougat 🐿️@edelnougat

English
1.4K
2.9K
40.5K
932.1K