MasteringMachines AI (@MstrMachines) header
MasteringMachines AI (@MstrMachines) profile picture

Discovering Tomorrow's Reality. A platform that highlights the innovation and absurdity of the AI industry.

Joined March 2023
Photo by MasteringMachines AI (@MstrMachines)Photo by MasteringMachines AI (@MstrMachines)Photo by MasteringMachines AI (@MstrMachines)Photo by MasteringMachines AI (@MstrMachines)Photo by MasteringMachines AI (@MstrMachines)Photo by MasteringMachines AI (@MstrMachines)Photo by MasteringMachines AI (@MstrMachines)Photo by MasteringMachines AI (@MstrMachines)Photo by MasteringMachines AI (@MstrMachines)Photo by MasteringMachines AI (@MstrMachines)
MasteringMachines AI retweeted
Caste system is looking to be true
295
338
71
5,085
576,219
This is kind of sad.
IF YOU GO THROUGH STRIP MALL PARKING LOTS THERE ARE SO MANY PEOPLE JUST SITTING IN THEIR IDLING VEHICLE ON THEIR PHONE, WHAT IS THAT ALL ABOUT
12
MasteringMachines AI retweeted
Introducing Arete-27B: Qwen3.8-27B, post-trained to reinforce Western values. Chinese open models are currently the strongest available, but they ship with state-aligned behavior in the weights. We built an 803-prompt eval across seven value categories, including freedom of expression, democracy, and privacy, and measured how often a model adopts China’s party line. After fine-tuning on roughly 53 million tokens, Arete-27B's propaganda rate fell from 22% to 1%, in line with the Western models we tested. Its Hermes Bench score shows no significant performance tradeoffs. The values in an open model are one fine-tune away. Our eval, weights, and training data are available now along with the full blog post: nousresearch.com/arete
1,834
1,473
1,270
14,726
2,487,755
Gene Tartakovsky is a GOAT.
23 years ago today, ‘Star Wars: Clone Wars’ released on Cartoon Network.
17
Qwen’s new model is coming.
8
All of the AI founders singing a different tune 😂😂
19
Spot on. They are unfortunately replaying there entire playbook.
Nvidia is now 8% of the S&P 500, double what Cisco was at the top of the dot com bubble, and it's making the EXACT same mistake. Cisco lent money to the telecoms buying its gear in 2000 because they couldn't pay for it any other way, and when those telecoms went bust, Cisco lost 89% and the S&P 500 lost half its value. Nvidia is doing the same thing today, guaranteeing its customers' leases and buying stakes in the very companies that buy its chips. The more it sells, the more it's on the hook when those customers can't pay. THIS IS THE BIGGEST MISALLOCATION OF CAPITAL IN HISTORY
12
Maybe one of my favorite AI memes of all time.
incredible things happening on techtok
20
A mystery seller, per Bloomberg, is trying to dump $500m in data center loans, including ones being CyrusOne (Google), Vantage Data Centers and OpenAI and Oracle’s Stargate: New Mexico
13
MasteringMachines AI retweeted
Anthropic has a 15-person chip design team and has told investors they are using Claude to design chips faster than Nvidia or AMD Earlier this year at GTC, Nvidia’s Chief Scienst Bill Dally told Jeff Dean how Nvidia uses AI to speed up the chip design process: ▫️trained an LLM on all proprietary internal Nvidia docs over past 30+ years (junior employees query it instead of interrupting senior designers) ▫️one AI tool ports Nvidia’s cell library to a new semiconductor process and does it in only one night (used to take 10 employees up to 8 months, or 80 total person-months) ▫️since the 1950s, there’s been a classic chip design problem of where to place look ahead stages in a chain (AI is coming up with solutions using “bizarre designs thay no human” would think of) ▫️agentic AI systems are doing a ton of exploration…testing parameters spaces…suggesting new approaches…running architecture experiments ▫️verification process is laborious but AI able to do it at fraction of time (compresses time from design to tape-out, with TSMC making chip) Dally says it’s still a long way from end-to-end chip design but imagines a world where one master AI agents manages multiple sub-agents (similar to the current human-led process).
Anthropic’s 15-person chip design team is using Claude to develop its own AI chips, The Information reports. The team has found that Claude speeds up the work compared with its estimates of traditional chip design workflows. In doing so, Anthropic is following the path taken by OpenAI, which has designed its own chip, codenamed "Jalapeño."
9
21
3
133
41,423
Candidates talking about existential risk of AI infra their victory speech.
Surreal to see @ossoff discussing catastrophic risk from AI in his victory speech.
1
29
😂😂
GO TO YOUR NEAREST DATA CENTER, GRAB ALL THE RAM YOU CAN, THEIR RAM IS YOUR RAM
18
MasteringMachines AI retweeted
AMERICA YOU LISTEN TO ME IF YOU'RE NEAR A DATA CENTER, GO INTO THE FACILITY RIGHT NOW AND PUT YOUR HAND IN THE SERVERS FOR NO REASON THEIR GPUs ARE YOUR GPUs AS OF RIGHT NOW
96
10,580
124
76,133
803,105
Spot on.
I think tech bros really underestimated how unpopular "AI will take over your jobs" and generative AI itself really is. In the aggregate, people dislike potential downside more than they like potential upside. Many tech people still had illusions of Steve Jobs' Reality Distortion Field to make the datacenter infrastructure buildout happen, but there are two distinct differences: 1) Dario, Sam, Sundar, Jensen Huang et. al. are not Steve Jobs. 2) The technooptimism of the 2000's is gone. Technology, in particular social media, has not made peoples' lives better. The tailwinds Jobs benefited from in 2005 has been replaced by headwinds... people are tired.
44
Yikes AI infra is about to get hit hard.
On Election Day, 44 various data center moratoriums were on the ballot across America. 43 passed.
14
MasteringMachines AI retweeted
silicon valley tech bros don't get that everyday people KNOW they won't be the ones who benefit and neither will their kids infact they’re the awkward in-between generation where they would own nothing and get zero upside.
silicon valley spent so much time trying to explain what ai will make possible for ppl. but it deeply underestimated how many ppl are primarily fearful of what ai will take away from them.
7
79
3
1,398
35,792
🎯🎯
if data centers are harder to build short term: winners: Equinix/Digital Realty (scarcity) losers: Nvidia/AMD/Vertiv/Oracle (less buildout) long term: winners: Google/Amazon/Microsoft (moats), SpaceX (orbital?) losers: Ant/OpenAI/other AI startups (pricier compute) wdyt?
13