Matt

@SawyerMerritt I don’t understand why US doesn’t sell model YL ? People need a real family car. Model Y is really too small for family. Cybertruck is not for family. Cyber SUV is not in any product roadmap. Sigh … , all the focus is on robots, cars seems not the priority for Tesla anymore.

If Tesla offered this SUV lineup in the US: • Model Y (188" long, two rows) • Model Y L (196" long, three rows) • Cyber SUV (~210" length, same as GMC Yukon, three rows) This would be one hell of an all-electric self-driving capable SUV lineup that would serve the needs of families of pretty much all sizes.

LLM Knowledge Bases Something I'm finding very useful recently: using LLMs to build personal knowledge bases for various topics of research interest. In this way, a large fraction of my recent token throughput is going less into manipulating code, and more into manipulating knowledge (stored as markdown and images). The latest LLMs are quite good at it. So: Data ingest: I index source documents (articles, papers, repos, datasets, images, etc.) into a raw/ directory, then I use an LLM to incrementally "compile" a wiki, which is just a collection of .md files in a directory structure. The wiki includes summaries of all the data in raw/, backlinks, and then it categorizes data into concepts, writes articles for them, and links them all. To convert web articles into .md files I like to use the Obsidian Web Clipper extension, and then I also use a hotkey to download all the related images to local so that my LLM can easily reference them. IDE: I use Obsidian as the IDE "frontend" where I can view the raw data, the the compiled wiki, and the derived visualizations. Important to note that the LLM writes and maintains all of the data of the wiki, I rarely touch it directly. I've played with a few Obsidian plugins to render and view data in other ways (e.g. Marp for slides). Q&A: Where things get interesting is that once your wiki is big enough (e.g. mine on some recent research is ~100 articles and ~400K words), you can ask your LLM agent all kinds of complex questions against the wiki, and it will go off, research the answers, etc. I thought I had to reach for fancy RAG, but the LLM has been pretty good about auto-maintaining index files and brief summaries of all the documents and it reads all the important related data fairly easily at this ~small scale. Output: Instead of getting answers in text/terminal, I like to have it render markdown files for me, or slide shows (Marp format), or matplotlib images, all of which I then view again in Obsidian. You can imagine many other visual output formats depending on the query. Often, I end up "filing" the outputs back into the wiki to enhance it for further queries. So my own explorations and queries always "add up" in the knowledge base. Linting: I've run some LLM "health checks" over the wiki to e.g. find inconsistent data, impute missing data (with web searchers), find interesting connections for new article candidates, etc., to incrementally clean up the wiki and enhance its overall data integrity. The LLMs are quite good at suggesting further questions to ask and look into. Extra tools: I find myself developing additional tools to process the data, e.g. I vibe coded a small and naive search engine over the wiki, which I both use directly (in a web ui), but more often I want to hand it off to an LLM via CLI as a tool for larger queries. Further explorations: As the repo grows, the natural desire is to also think about synthetic data generation + finetuning to have your LLM "know" the data in its weights instead of just context windows. TLDR: raw data from a given number of sources is collected, then compiled by an LLM into a .md wiki, then operated on by various CLIs by the LLM to do Q&A and to incrementally enhance the wiki, and all of it viewable in Obsidian. You rarely ever write or edit the wiki manually, it's the domain of the LLM. I think there is room here for an incredible new product instead of a hacky collection of scripts.

This is a crazy clip

@TeslaBoomerMama Amazing to foresee this before the announcement

Step 1 ✅ Hopefully my article is right on the other details, too

You heard it today, the countdown for FSD approval has started! It's time for TESLA OWNERS and TSLA RETAIL to act! Please take the time to help! Following this link you will be able to access names & email addresses and - if desired - a template to contact your local legislators / regulators in different countries. t.co/XlRsZWoW2c You can use either the contact database and a static version only, or if you want the full experience, hit the "dynamic email generator" and thereafter the "copy" button, you then have your independent set up. Then use the tabs on the bottom of the newly generated Google sheet to navigate to your desired email outcome. In any case, nobody here sees what you are doing and you are not sending any information to us. Thank you for your help with getting FSD approved. And thanks to @_jeffreyr who kindly set this up

@SawyerMerritt This is very promising for all EV drivers, chargers everywhere

Another business has purchased Tesla Superchargers and put their own logo on them. 16 charging stalls.

@DavidMoss You made the history!

@DavidMoss incredible journey !

The record setting Tesla FSD intervention streak is over We made it so close to 13k at 12,961 miles (20,858KM) but rural Wisconsin roads + snow + weather in the teens was just a bit too much. No accident or anything big to note just found myself in a tricky situation where it was time to takeover. Safety over everything people! I will be posting the footage later today from the saved Dashcam clip. FSD 14.2 took me to 30 states & 7.5 weeks of driving without any human input whatsoever. That’s absolutely every part of the driving end to end for weeks on end. All I did was put in the navigation destination & press start self driving & it always took me to where I wanted to go.Pretty incredible & that’s before even talking about me being the first person to drive coast to coast fully autonomously with no interventions. All the support you guys have showed these past 2 months have meant the absolute world to me, I wish I could have the chance to thank you all in person someday. I appreciate it all very much. We aren’t going anywhere people, this streak may be over but the journey is just getting started.

$TSLA #FSD Massive shift in the last 6 months

First Cross Country Autonomous Drive Completed! Over 10,000 Miles Using Tesla FSD w/ No Intervention youtu.be/dnLswbNB0SU?si…

@aelluswamy Just like that !

And so it begins!

@HansCNelson @Tesla Same feeling as you said : “I don't care how fast an ICE car can fuel up. I NEVER want to drive that far without @tesla FSD again.”

Had my first LOOONG road trip experience on V14 over the past week. It drove over 99.9% of all the miles for the whole trip. Holy smokes. It was so good! I don't care how fast an ICE car can fuel up. I NEVER want to drive that far without @tesla FSD again.

@ItsKimJava @Tesla Upstate NY 3 days trip. FSD v14.2 is amazing !

Finally tried @Tesla’s Full Self Driving v14.2 and it’s legitimately scary good. Holiday packed parking lots and construction zones with incredible confidence and prediction. The way it just… gets it. Feels less like software and more like the car borrowed a human brain for the day. Speechless.

@SawyerMerritt My 3 days trip to upstate NY, 99% FSD, amazing

This guy’s Cybertruck just drove itself 1,200 miles from Chicago to Cape Canaveral, Florida without him touching the wheel or pedals a single time. How is everyone not talking about this?! Self-driving cars aren’t some far-off thing, they’re already here!

The @ilyasut episode 0:00:00 – Explaining model jaggedness 0:09:39 - Emotions and value functions 0:18:49 – What are we scaling? 0:25:13 – Why humans generalize better than models 0:35:45 – Straight-shotting superintelligence 0:46:47 – SSI’s model will learn from deployment 0:55:07 – Alignment 1:18:13 – “We are squarely an age of research company” 1:29:23 – Self-play and multi-agent 1:32:42 – Research taste Look up Dwarkesh Podcast on YouTube, Apple Podcasts, or Spotify. Enjoy!

$TSLA Tesla FSD is so good now , from start to destination for one hour , no human intervention for the whole trip , highway plus local traffic , flawless. Amazing !

- the yen's weakening alongside rising Japanese bond yields to potential Chinese capital outflows from Japanese assets, linking it to Prime Minister Sanae Takaichi's recent Taiwan comments. - Takaichi's November 7 remarks in parliament hinted at Japanese military support for Taiwan against Chinese aggression, breaking from strategic ambiguity and sparking a diplomatic row, as covered by The Guardian and CGTN. - The ensuing Japan-China tensions have driven a market selloff, pushing 10-year JGB yields to 1.73% and 20-year yields to 2.815%—the highest since 1999—while tourism and retail stocks plunged, according to Bloomberg and WSJ analyses.

As illustrated in these Bloomberg charts, the Japanese Yen has been weakening toward year-to-date levels despite rising government bond yields and a narrowing spread with the US. Look for the government to step up its “oral intervention,” fueled by several economic concerns.

@KobeissiLetter Too much MSG and too salty

Either Chipotle has collapsed as a brand or young Americans are out of money. Or, both. What is happening here?

M3+DGX