Pinned Tweet
Tesla Is A Monopoly, We Just Don’t Know It Yet by Corey Recvlohe link.medium.com/8fLgxsyZrX
$TSLA #tesla
NVIDIA’s NOOA Proves the Harness Matters More Than the Model, and Everyone’s Watching the Wrong… by @glenn_all pub.towardsai.net/nvidias-no…
1
42
Miguel P3ra1v retweeted
Replying to @pwendell @databricks
Full post here:
databricks.com/blog/managing…
Miguel P3ra1v retweeted
Today @databricks we're publishing a detailed analysis of techniques we used to drastically reduce our internal AI spend while aggressively growing adoption. Savings come from layering in several techniques, which combine to drive unit costs down as much as 90% in some scenarios. Tl;dr, the wins come from:
1. Shifting defaults to more efficient models, including OSS models such as GLM. Maximum intelligence models simply aren't needed for many coding tasks, and "good enough" models are quickly becoming very cheap. We shift traffic between models using Unity AI Gateway. Approximate savings: 50% or more.
2. Using smart routing to automate model selection. Routing can further squeeze efficiency by dynamically selecting the model or harness that can most efficiently execute a particular task. Our task-level routing leverages @omnigent_ai. Approximate savings: 30%.
3. Providing user visibility and adaptive budgeting. Every user can see how much they spend, and users receive hints on how to contain spend. Heavy spenders encounter progressive friction as they ratchet spend above certain levels. Approximate savings: 10%.
4. Managing context bloat by pruning tool call results and tuning harness settings. Extraneous context costs $$ and delivers no value. Tuning cache settings also help lower average token costs. Approximate savings: 10%.
Miguel P3ra1v retweeted
A few random notes from claude coding quite a bit last few weeks.
Coding workflow. Given the latest lift in LLM coding capability, like many others I rapidly went from about 80% manual+autocomplete coding and 20% agents in November to 80% agent coding and 20% edits+touchups in December. i.e. I really am mostly programming in English now, a bit sheepishly telling the LLM what code to write... in words. It hurts the ego a bit but the power to operate over software in large "code actions" is just too net useful, especially once you adapt to it, configure it, learn to use it, and wrap your head around what it can and cannot do. This is easily the biggest change to my basic coding workflow in ~2 decades of programming and it happened over the course of a few weeks. I'd expect something similar to be happening to well into double digit percent of engineers out there, while the awareness of it in the general population feels well into low single digit percent.
IDEs/agent swarms/fallability. Both the "no need for IDE anymore" hype and the "agent swarm" hype is imo too much for right now. The models definitely still make mistakes and if you have any code you actually care about I would watch them like a hawk, in a nice large IDE on the side. The mistakes have changed a lot - they are not simple syntax errors anymore, they are subtle conceptual errors that a slightly sloppy, hasty junior dev might do. The most common category is that the models make wrong assumptions on your behalf and just run along with them without checking. They also don't manage their confusion, they don't seek clarifications, they don't surface inconsistencies, they don't present tradeoffs, they don't push back when they should, and they are still a little too sycophantic. Things get better in plan mode, but there is some need for a lightweight inline plan mode. They also really like to overcomplicate code and APIs, they bloat abstractions, they don't clean up dead code after themselves, etc. They will implement an inefficient, bloated, brittle construction over 1000 lines of code and it's up to you to be like "umm couldn't you just do this instead?" and they will be like "of course!" and immediately cut it down to 100 lines. They still sometimes change/remove comments and code they don't like or don't sufficiently understand as side effects, even if it is orthogonal to the task at hand. All of this happens despite a few simple attempts to fix it via instructions in CLAUDE . md. Despite all these issues, it is still a net huge improvement and it's very difficult to imagine going back to manual coding. TLDR everyone has their developing flow, my current is a small few CC sessions on the left in ghostty windows/tabs and an IDE on the right for viewing the code + manual edits.
Tenacity. It's so interesting to watch an agent relentlessly work at something. They never get tired, they never get demoralized, they just keep going and trying things where a person would have given up long ago to fight another day. It's a "feel the AGI" moment to watch it struggle with something for a long time just to come out victorious 30 minutes later. You realize that stamina is a core bottleneck to work and that with LLMs in hand it has been dramatically increased.
Speedups. It's not clear how to measure the "speedup" of LLM assistance. Certainly I feel net way faster at what I was going to do, but the main effect is that I do a lot more than I was going to do because 1) I can code up all kinds of things that just wouldn't have been worth coding before and 2) I can approach code that I couldn't work on before because of knowledge/skill issue. So certainly it's speedup, but it's possibly a lot more an expansion.
Leverage. LLMs are exceptionally good at looping until they meet specific goals and this is where most of the "feel the AGI" magic is to be found. Don't tell it what to do, give it success criteria and watch it go. Get it to write tests first and then pass them. Put it in the loop with a browser MCP. Write the naive algorithm that is very likely correct first, then ask it to optimize it while preserving correctness. Change your approach from imperative to declarative to get the agents looping longer and gain leverage.
Fun. I didn't anticipate that with agents programming feels *more* fun because a lot of the fill in the blanks drudgery is removed and what remains is the creative part. I also feel less blocked/stuck (which is not fun) and I experience a lot more courage because there's almost always a way to work hand in hand with it to make some positive progress. I have seen the opposite sentiment from other people too; LLM coding will split up engineers based on those who primarily liked coding and those who primarily liked building.
Atrophy. I've already noticed that I am slowly starting to atrophy my ability to write code manually. Generation (writing code) and discrimination (reading code) are different capabilities in the brain. Largely due to all the little mostly syntactic details involved in programming, you can review code just fine even if you struggle to write it.
Slopacolypse. I am bracing for 2026 as the year of the slopacolypse across all of github, substack, arxiv, X/instagram, and generally all digital media. We're also going to see a lot more AI hype productivity theater (is that even possible?), on the side of actual, real improvements.
Questions. A few of the questions on my mind:
- What happens to the "10X engineer" - the ratio of productivity between the mean and the max engineer? It's quite possible that this grows *a lot*.
- Armed with LLMs, do generalists increasingly outperform specialists? LLMs are a lot better at fill in the blanks (the micro) than grand strategy (the macro).
- What does LLM coding feel like in the future? Is it like playing StarCraft? Playing Factorio? Playing music?
- How much of society is bottlenecked by digital knowledge work?
TLDR Where does this leave us? LLM agent capabilities (Claude & Codex especially) have crossed some kind of threshold of coherence around December 2025 and caused a phase shift in software engineering and closely related. The intelligence part suddenly feels quite a bit ahead of all the rest of it - integrations (tools, knowledge), the necessity for new organizational workflows, processes, diffusion more generally. 2026 is going to be a high energy year as the industry metabolizes the new capability.
Miguel P3ra1v retweeted
I'm not joking and this isn't funny. We have been trying to build distributed agent orchestrators at Google since last year. There are various options, not everyone is aligned... I gave Claude Code a description of the problem, it generated what we built last year in an hour.
Miguel P3ra1v retweeted
🇺🇸 SAN FRAN BLACKOUT – WAYMO FROZE, TESLA DROVE
Waymo’s robotaxis got a little too real last night - by completely shutting down when San Francisco’s power outage knocked out traffic lights.
Meanwhile, Teslas on FSD? Kept rolling. No drama, no headlines - just handling chaos like it’s a walk in the park.
This is what happens when you train your AI on billions of real-world miles instead of coddling it in a simulation padded with perfect data and wishful thinking.
Waymo bet on maps and order. Tesla bet on mess - and won. When the lights go out, the difference isn’t theoretical. It’s traffic.
Source: @teslasi, @elonmusk
Miguel P3ra1v retweeted
Revolutionary Magnetic Seal Stops Water, Acid, and Oil Leaks Instantly
by @IntEngineering
#Innovation #EmergingTech #Technology #Tech
cc: @karpathy @terenceleungsf @kuriharan
Miguel P3ra1v retweeted
With over $280 billion of total assets under management and over $1 billion in Tesla shares, the State Board of Administration of Florida Retirement System came out strongly in favor of our incentive proposal for Elon, designed to drive extraordinary growth, which the SBA describes as the “gold standard”
See the SBA’s full comments below
sec.gov/Archives/edgar/data/…
Miguel P3ra1v retweeted
Tesla FSD 14.1.3 reverses at a construction zone to allow pedestrians to pass. Amazing! @teslasi @aelluswamy @wholemars @hletes
Miguel P3ra1v retweeted
Just did a 4 hour drive this morning testing FSD V14.1.4.
It was 100km (62 Miles) of driving all throughout multiple cities, including two costcos and countless destination parkings.
I did not have to intervene once the entire time.
THAT IS SO CRAZY - we are living in the future.
I look out at all of the other cars on the roads and I truly do not think people understand what is about to happen. FSD is going to change the world.
Incredible work @teslasi
Miguel P3ra1v retweeted
🚨🚨 Morgan Stanley's Adam Jonas officially says his money is on Tesla in the autonomy race:
"I'm callin' it. Autonomous cars are solved. Do I mean six or seven 9's to the right of the decimal? No. Perfection? Never. But enough to pull the safety driver at scale in major metros. From our discussions with Tesla, it appears the only thing preventing Tesla from pulling the driver is its own abundance of caution."
Morgan Stanley's Adam Jonas in new $TSLA note: "I'm callin' it. Autonomous cars are solved. Do I mean six or seven 9's to the right of the decimal? No. Perfection? Never. But enough to pull the safety driver at scale in major metros."
"From our discussions with @Tesla, it appears the only thing preventing Tesla from pulling the driver is its own abundance of caution. Our understanding is that there is no other explicit regulatory approval required to pull drivers in Texas. A passive optical-only AV system would seriously challenge the conventional thinking of many in the robotaxi community. Tesla truly believes that radar and LIDAR do not make its cars any safer. It is our longstanding opinion that if you solve autonomy for cars – you solve autonomy for many other form factors of AI-enabled robotics (aviation, marine, weapons, etc).
Tesla will have around 8 million vehicles in the global 'parc' by year-end. 12% FSD penetration (paying monthly subs and upfront purchasers) implies 1mm active paying users. If we were to apply the $99/month subscription fee to all users (including upfront payers for comparative equivalency) implies approx $300mm of quarterly FSD revenue or $1.2bn annualized FSD revenue at a year-end exit rate. Assuming an 85% margin would imply approx $1bn of EBIT... accounting for 15 to 20% of total company EBIT. Highly profitable, highly reoccurring."
Miguel P3ra1v retweeted
It’s the little things
FSD 14 can recognize people’s gestures indicating they’re about to leave a busy parking lot at Costco.
This prompts FSD to halt and wait until the car departs and parks seamlessly in the exact spot. What a generational update.
Good job @Tesla
Miguel P3ra1v retweeted
First impression of FSD v14
It simply feels more precise and intentional than v13.
Kind of like driving a sport-tuned car after having driven a squishy/drifty luxury sedan.
It is punchier and whippier in ways that are confidence inspiring (on windy curves alongside traffic, in lane shifts etc.)
And then it pulls you into the parking spot.
🦾
Miguel P3ra1v retweeted
We build world models grounded in physical reality, naturally generalizing across cars and robots 🤖
Miguel P3ra1v retweeted
Wrote about some of the recent work from the excellent @teslasi team. The same material was presented at the International Conference of Computer Vision this week.
The main purpose of sharing this content is to show the cutting-edge work that the team has been developing and attract more talented engineers to join the team.
Tesla is THE place to solve real-world robotics. The team is extremely talented, determined and motivated. The technology that is already developed is cutting-edge and the roadmap is exhilarating. The work done here will tremendously benefit all of humanity. Hence, we believe that this is the best place to work on AI on the entire planet currently. Join us and let's bring millions of intelligent, friendly and useful robots to life.
tesla.com/AI























