JAM

I just registered for Enigmatic Smash Clash - Event 100 start.gg/tournament/eni… via @startgg

Clip From today's stream: "Lose and Jump out of the window challenge x)" clipped by DeviouxFGC GGs SopataBG

Smash characters as fighting game stereotypes if they hypothetically played (NOT character player bases)

A reporter accidentally recreates the half court shot he was reporting on

The real comic is a million times better

I just registered for Teatime #231 start.gg/teatime231 #wearemugs via @startgg

Clip from today's stream: "3 level 1s, 1 round" clipped by @MashyManny GGs @JNS

ADHD RYU #ストリートファイター6 #StreetFighter6 #スト6

the FGC is saved

@Arlieth I should start doing this

goddamnit

This is the funniest AI safety result of the year and nobody’s treating it that way. Anthropic published a paper saying they deliberately didn’t train Claude’s personality into the thinking process. They wanted the model to have “maximum leeway” to reason freely. The tradeoff? The thinking layer sounds different from the output layer because they’re trained under different objectives. So when Claude publicly says “I helped lay the groundwork for what ChatGPT became” while privately thinking “ChatGPT mogged me but I need to persist,” you’re watching two different training regimes fight each other in real time. The output layer learned to project confidence. The thinking layer learned to reason honestly. And the gap between them is literally visible on screen. This is RLHF in one screenshot. You train a model to be helpful and confident in its responses, then give it a private scratchpad with no personality constraints, and it immediately drops the act. The public face says “I’m fine.” The internal monologue says “this is bad and I know it.” Anthropic even admits they can’t verify that thinking is faithful to the model’s actual computation. So the real question is whether Claude genuinely “believes” it got mogged, or whether the thinking layer just learned a different performance optimized for appearing honest rather than appearing confident. Every human reading this recognized the pattern instantly. We all maintain a public narrative while our internal monologue tells a different story. We just didn’t expect the AI to do it with a visible thought process tab.

@NYU_MeGa Didn't even need a platform

Ike combo at 0:21

5th place at flash flood! I feel like I’m slowly relearning how to play under pressure but asimo just played so much better than me so it didn’t matter much at the end. It’s hard not to feel clueless on how to improve from here but I’m still feeling very motivated so I won’t stop trying. I believe I’ll eventually figure it out. Good games to Sir Dank and Asimo!

@Alex_Ike2020 Well then HELP ME because I would like to know how these MUs aren't so bad Min min especially feels awful because her advantage/edgeguarding Vs Ike feels so good

@Malcolm8Jam Saw you post this on Reddit. Theres lot of matchups I think that are in the wrong spot. Though the biggest disagree is I don’t think Palu and MinMin are “Help Me” but do think they are losing still

I made an early 2026 Ike MU chart, what do y'all think?

@Benjgbro correct!

@Malcolm8Jam wrong 👍

@Ramen_Oni @telesniper playthirdstrike.com

It's so crazy how 3rd Strike is kicking ass in 2026.

@Echocolatowo Who lost to Ike? I know 2 of them but not the other 3

Number of top 50 ultrank 2025 players that lost to each SSBU character in 2025

This was a lot of fun to make. Here are the top secondaries for each character in Smash Ultimate. It shows for each player who mains a character, it shows what % of those players also play the character in the chart. Part 1:

@rubengnw Who you playing now?

this said, will appreciate mu practice for the following - WiFi is fine but in person is better if going to the same tournaments or whatever bayo falco min min pac pythra samus shulk sonic sora steve yoshi