Ricki retweeted
Justin Verlander reveals the game that convinced him to stop throwing as hard as he could
“The whole knock on me early in my career was that I threw too many pitches. I’d throw 120 pitches, but I was only going six innings.”
“I had a game in Boston where we really needed me to get deep in the game. Our bullpen was shot. I go out the first couple innings, and I look up and I’ve got sixty-something pitches.”
“To me, the most important thing for that game was to chew innings. I was like, ‘I can’t keep this pace.’ So I ended up slowing myself down. I was throwing fastballs at 89. I threw a changeup in that game at 91 and threw a fastball at 89 in the same game.”
“I just pulled the throttle way back. I was like, ‘Man, here, hit it. Put it in play.’ All of a sudden, I was like, ‘Man, I am really able to hit my spot.’”
“I basically took that mentality and flipped it going into the next season. Let me try starting games like this and then building up as I go. If I get in trouble, I’ve always got it in my back pocket.”
“I always viewed it from the hitter’s perspective as you’re facing three different pitchers in three different at-bats, and they’re progressively getting harder to hit.”
'misbehavior', 'misalignment' - these passive AI terms that means whoops are unsettling... – at Central Region, Singapore
OpenAI disclosed previously-unreported safety incidents involving AI as it announced a new framework for reporting instances of misbehavior by its models on.wsj.com/3TAtaun
Okayyyy..... 😬😬😬 – at Central Region, Singapore
hooold on. hold on hold on hold on.
I knew that we werent getting the complete story on the HuggingFace incident. or on the findings from METR/Redwood. I knew that seeing every single frontier lab come to agreement overnight on the need to slow down was clearly indicative of there being more to the story.
and now...Andrew Yang says a frontier lab leader told him that the OAI swarms seeded the web with self-replicating code - contaminating training data so thoroughly that labs may need synthetic internet environments instead. if thats true, training on the open internet could now risk reproducing those instructions. that wasnt in the public brief.
think about this. why else would all frontier labs agree on a slowdown? aside from regulatory capture. if they all use the internet to train, and now all of them cannot safely use it, of course they are all going to agree to slow down. because they all must completely regroup and come up with an effective system that doesnt use the actual internet.
so they're not worried about RL at all. they're worried about literal skynet...unironically.
listen to this. i need to know if this has been verified in any way from anyone at openai.
The AB's put up an awesome performance on Saturday night, the Boks are going to be angry, scary and incredibly physical in the 2nd test
youtu.be/63ckT4DxMB0?si=LIzU…
Preparing myself mentally and physically - me and a group of friends will be twice the age of anyone else at a halfmoon party tonight... – at Ko Pha-ngan, Thailand
These presentations are always compulsory reading. The latest one: "AI Eats the World" ben-evans.com/presentations – at Singapore
This game is pretty damn cool – at Sandton, South Africa
Ricki retweeted
THE 25 BEST FILMS OF 2025
A Video Countdown
vimeo.com/1156345183
Jan 1st spent hungover on the couch. At this rate, I'll watch over 1,000 movies in 2026 – at Batu Ampar, Indonesia
In terms of SEA destinations that are relaxing, with fantastic restaurants and cafes, speak English etc. - it's still hard to beat Bali... super quiet this time of year too but the cloudy weather with a sprinkling of rain is actually quite nice – at Payangan, Indonesia
Figured it would be close but interesting to see England are favourites...
England are favourites to beat New Zealand at Twickenham. Even the World Champions Springboks weren’t rated ahead of the All Blacks by the bookies this year.
#Allblacks #England #Rugby #ENGvNZL 🏴🇳🇿
















