Sergeant Patel

I don't have 30k points to get a cyberhammer so I just 3d printed one in metal using Desktop Metal Studio System. #tesla #cyberhammer #Cybertruck #3Dprinting

@SawyerMerritt Potholes of Massachusetts Changing lanes at split lane left turn with heavy traffic We should be able to prevent rerouting to save 1 minute instead build a logic to be able to increase time saving as custom if/else logic. If time save >=X minutes , then re-route

If you use any version of FSD (Supervised) V14, what is the most common reason you disengage (even if it’s rare)?

@RobinhoodApp Just the phone

Share your trading setup and we just might send you some swag.

Tesla has invented invisible Robotaxi supervisors! Regulators: “You must have a supervisor in the car.” Tesla: “We do.” Regulators: “Where?” Tesla: “Front left seat.” Regulators: “There’s no one there.” Tesla: “Exactly. Zero distractions. Perfect attendance. Never blinks.” Regulators: “That’s… not how supervision works.” Tesla: “Oh it gets better - she hasn’t had a single safety incident.” Regulators: “Because she doesn’t exist.” Tesla: “Or… because she’s flawless.”

@StockMKTNewz Robinhood Gold 3% cashback (most purchase) Citi double cash 2% cashback Citi custom cash 5% cashback grocery Citi custom cash 2, 5% cashback gas Discover rotating 5% category >$1500 cashback a year gauranteed

What credit cards do you have/use?

@Mr_Derivatives They saw people calling it, Strait of Trump.

Incoming Dad Joke…. I dare anyone who plays poker, to stand up when they go all in with an open ended straight draw on the flop and completes their straight on the last card, to yell out “my Strait of Hermuz, caught it on the river!” Har har har. Ok I’ll see myself out now.

@James95113983 @aakashgupta Because Elon posted about it

@aakashgupta Did you hear FSD influencers telling everyone this new Tesla AI chip is NOT for FSD, but for Tesla Optimus bot and for Tesla data centers?🙂

Tesla just taped out an AI chip that matches Nvidia's H100 on one specific workload. It can't train a language model. It can't run general ML research. It runs INT4, INT2, and FP8 integer math, and almost nothing else. That's the entire strategy. Nvidia's H100 is built to do everything. FP16 training, FP32 science, transformer attention, sparse operations. Every one of those capabilities costs die area, power, and cooling. The chip ends up drawing 700 watts and 80GB of HBM because it has to serve every customer in AI, from OpenAI to a 3-person startup. Tesla has one customer: a car. A car needs one thing: low-precision inference at the edge, 30 times per second, forever. Nothing else. So they stripped the chip down to that. No FP16 training. No HBM. No general-purpose scheduler. AI5 targets INT4, INT2, and FP8, the exact precisions autonomous driving models use after quantization. Everything else got cut. Tesla claims a single AI5 matches H100 performance on their workload. A dual setup rivals Blackwell. At a fraction of the cost and power. The full AI5 computer targets 2,000-2,500 TOPS vs 300-500 for AI4. 5-8x more compute on the exact math a car needs, with zero capability for anything else. This is the Google TPU playbook and the Amazon Trainium playbook. General-purpose chips have to be good at everything. Specialized chips only have to be good at the one workload that pays for them. If that workload ships in millions of units a year, the math works. Nvidia's business is selling one chip to every AI customer. Tesla's business is selling cars, and they just took Nvidia out of their own bill of materials. Fabbed at TSMC Arizona and Samsung Texas. The chip Tesla's entire autonomy roadmap depends on no longer routes through the Taiwan Strait.

@OnlyInBOS Shah's halal Worcester

If someone asked you for one halal food spot in Massachusetts, where are you sending them?

Yes I agree. My openclaw works better with claude. GPT 5.4 unable to meet the agentic demands. Just tried GLM 5.1 cloud hosted by ollama, slightly better but not quite the Claude. I cannot spend more than $100 a month so my openclaw is going to be mediocre until something changes.

I tried really hard to get GPT 5.4 to work with OpenClaw but it’s just not there. Opus 4.6 is just so much better. I wish there was better competition at the frontier. IMO it’s Anthropic, then everyone else at the moment. Grok is better at research and fact checking. Gemini and grok are both far better at image and video, mainly because Anthropic doesn’t do it. GPT 5.4 is very good for an overall good model for a good price that’s a generalist. But when it comes to specialization, OpenAI doesn’t have anything IMO. I think this is going to bite them really hard. However once players start leveraging their new compute clusters I think it’ll flip very quickly. Anthropic has made some pretty big errors as it pertains to securing compute in the long term it seems. If I were a betting man, I think Anthropic secure its lead by middle of this year, but then Grok and Gemini start making a huge comeback for 1st. Only path for OpenAI long term IMO is to ultra-specialize around openclaw and becoming the best bang-for-your-buck brain for agentic tools and harnesses.

@GregNotSure @NinaDSchick There is probably not enough gpus to get everyone use it.

@NinaDSchick I'm starting to understand why AI people are being universally hated. I am starting to hate all of you too, and I use AI daily. I hope that this market crashes and burns, and recovers as just another computer tool.

Claude Mythos. Ten trillion parameters: the first model in this weight class. Estimated training cost: ten billion dollars. On the hardest coding test in the industry (SWE bench) it scores 94%. It found a security flaw in a system that had been running for 27 years, one that every human engineer and every automated check had missed. It found another bug that had survived five million test runs over 16 years. (It did so overnight.) It is so capable in cybersecurity that Anthropic will not release it to the public, instead it is launching Project Glasswing along with 100m in compute credits to help secure software. Only twelve partners currently have access: Amazon, Cisco, Apple, Google, Microsoft, NVIDIA, JPMorgan Chase, Crowdstrike, Palo Alto, AWS, The Linux Foundation, Broadcom. (I'm sure the Pentagon is on the line?) This is not a product launch: it is a controlled deployment of a system too powerful to distribute freely. Tell me this isn't (very expensive) AGI?

@m0rph3V @alojoh @BoBbyPleWniaK No he didn't. Spotting AJ troll

@sgtpatel @alojoh @BoBbyPleWniaK And he told them again to gtfo at $490

Another 150 point pullback. Last year it was 260. Unreal this stupid stock $tsla.

@alojoh @BoBbyPleWniaK You were telling people to get in and buy at $410 same time last year

@BoBbyPleWniaK Not reading our research is very very very expensive to everyone invested in TSLA. Don't waste your time on st*pid accounts. Peace

Impressive use of AI

@stocktalkweekly 😂

Artemis II flight window opens in 1 hour

@joeroganhq slowly killing themselves

North Korea's "elite" Specials Forces.

@nishkidoonoo @heyitsyashu @farzyness @elonmusk Brought to you by a diverse group of people. Black optimus looks the best tbh.

@heyitsyashu @farzyness @elonmusk “white pipo want slaves back so bad they made black robots” 😂😂😂

Did @elonmusk just Netflix Optimus ???

@jimhwilliams1 @DavidSacks @grok Elon is too busy building real things. He doesn't have time for this.

@DavidSacks @grok why was Elon not part of this?

I am honored and grateful to be appointed by President Trump to the President’s Council of Advisors on Science and Technology (PCAST) and to be named Co-Chair along with OSTP Director Michael Kratsios. PCAST is the principal body of external advisors tasked with shaping science, technology, and innovation policy for the President and the White House. Thirteen of the world’s most accomplished leaders in science and technology will join us as this PCAST’s initial members. Together we will make policy recommendations to ensure that America leads—and wins—in artificial intelligence and other cutting-edge technologies.  I look forward to working with the initial members: Marc Andreessen, Sergey Brin, Safra Catz, Michael Dell, Jacob DeWitte, Fred Ehrsam, Larry Ellison, David Friedberg, Jensen Huang, John Martinis, Bob Mumgaard, Lisa Su, and Mark Zuckerberg. Thank you to President Trump for his visionary leadership on technology policy which attracts the top luminaries in their fields to serve. It is an honor to be part of this distinguished group.

@pmarca I will send this book to my agent so he can read it in his living room.

Birth of the AI book. 🤖🤝🧑‍🏫 tylercowen.com/marginal-revol…

@mrfundman Buying with margin. Not selling tesla

Who’s investing in SpaceX IPO? Are you planning to sell any of your $TSLA shares or use new capital?

@farzyness That would be funny

Prediction: April 1st, Tesla will list a Cybercab for sale with steering wheel and pedals. People will think it’s fake. It won’t be. The greatest April Fools joke in human history.

@wholemars That's actually $25k per car for Uber. If they do end up getting 50k cars. Loss for Rivian in the long run if they can't figure out to make them cheaper

“We’ll buy 10,000 and then maybe 40,000 more if it actually works”