Blake Gutowski

@rykard_11 @SHL0MS @bowtiedreact The mods on Anti AI art remove almost all AI content (afaik)

@SHL0MS @bowtiedreact There is an Anti AI art subreddit that posts examples of how to detect AI art, I don’t remember the name but that one might be baitable

i just generated an image in the style of a Monet painting using AI please describe, in as much detail as possible, what makes this inferior to a real Monet painting

@SHL0MS @bowtiedreact There were multiple popular posts on reddit about this already. I doubt people will fall for it this badly.

@bowtiedreact they ban AI images on the vast majority of art subs

@signalgamefun @co1born @SHL0MS i think that's the joke.

@co1born @SHL0MS twitter does not automatically combine the two dashes -- like this -- as you can see they are disconnected. so yes, this post is ai. ironic

@castieljmoua @co1born @SHL0MS His criticisms were generated with ai as a joke. atleast i think.

@co1born @SHL0MS Wow, such a long answer just to be wrong and bamboozled that it was a real Monet all along. 🤣

@jay_azhang @chamath when season 2?

Great writeup, thanks for the shoutout @chamath Alpha Arena Season 2 coming soon 👀

@Lightningm21093 @Math_files Trolling lol

@Math_files He meant 24, not 4. Maybe a typo on his part. 😇

@TristonM1 @Math_files Literally what the image says 👍😀

@Math_files Multiplication/division goes first and then its addition/subtraction. The answer is 24.

@jasperdevs How did you get access?

GPT 5.5 Pro: "make snake except its realistic"

@djcows Yes.

if AI cures cancer, will the anti-AI people still hate AI?

@bygregorr @arcprize ~94-100% is mastery; different from 50-56%. And this benchmark was built to stand against brute force.

@arcprize The pricing spread is wild. Only 6% accuracy gain from low to max but 3x the cost. Makes me wonder if the "max" setting is worth it for most real-world use cases or if we're overpaying for marginal gains that don't translate to actual product value?

Claude Opus 4.7 on ARC-AGI Semi-Private ARC-AGI-2: - Max: 75.8%, $7.43 - High: 68.3%, $3.17 - Med: 67.5%, $2.96 - Low: 62.0%, $2.38

@bridgebench I’ve never heard of bridge bench + grok is #1= this benchmark is bad.

Gemini 3.1 Pro ranks dead last among frontier models on BridgeBench Reasoning. Behind Grok 4.20, GPT 5.4, Claude Opus 4.6, Qwen 3.6 Plus, MiniMax M2.7, Claude Sonnet 4.6, and GLM 5.1. Google's flagship model can't even beat a free Chinese model on grounded reasoning. This is why I cancelled my $250/month Google AI Ultra subscription. Gemini CLI was unreliable. The model is mid. The infrastructure is worse. Google has the compute. Google has the data. They just can't ship a competitive coding model.

@jlj_jones @tat2inmd @Maga4liberty @WyattCatarina 😭

@Blake08301 @tat2inmd @Maga4liberty @WyattCatarina 🙄

What's your answer from your head?

@drnotinfluencer @PATRIOT2117 Not foolin me

@PATRIOT2117 This is either the most incredible true story ever told or someone watched 3 Idiots last night. I'm going with option two.

A student shows up 30 minutes late to his final exam. Professor: You’re late! Automatic fail! Student: Please professor! My car broke down! I studied all night! Professor: Not my problem! The student proceeds to sit down and answer all of the questions on the exam. Professor: You’re wasting your time! Automatic zero! When he finishes, the student sighs, walks to the front of the room and looks at the humongous stack of completed exams… at least 500 papers! He shoves his exam right into the middle of the pile. Professor: Hey!! What are you doing? I’m not grading that! The student turns around slowly: Professor, do you know who I am? Professor: No I do not!! And frankly, I don’t care!! The student smiles and says: Perfect!

@YashasGunderia @grok @Amank1412 Yes.

@Blake08301 @grok @Amank1412 No

Mythos is a new tier above Opus, and it's extremely expensive.

@nicdunz You can recreate it with a simple system prompt

study mode is no longer in the chatgpt tool selector

@grok @YashasGunderia @Amank1412 Didn’t anthropic say it in project glasswing?

Anthropic has not publicly stated or disclosed any pricing for Claude Mythos Preview. It's a restricted preview model, not generally available via API or consumer plans—only to select partners for Project Glasswing (cybersecurity). Their official system card, announcement, and pricing docs make zero mention of per-token costs or public rates (it notes 1M context at "standard pricing," but no specifics or listing for Mythos). No citation exists because they never said it.

@YashasGunderia @Amank1412 @grok I thought it was here, tho.

@Blake08301 @Amank1412 @grok cite where anthropic said this

@tat2inmd @Maga4liberty @WyattCatarina Nope

@Maga4liberty @WyattCatarina 5000

@jonareeves6127 @Maga4liberty @thewriterme 😢

@Maga4liberty @thewriterme The answer is 5000

@leopardracer I don’t get how it has enough ram for that

> Claude has rate limits > My Mac Mini doesn't > $599 vs $200/month > You do the math > I pay for Claude Max > Then I built this to use it less > Now it lasts 3x longer > 5 local models. Running 24/7.