Sabitlenmiş Tweet
4 AI Breakthroughs Changing Robotics Forever: DrEureka, Voyager, Darwin-Gödel Machines & AlphaEvolve Explained x.com/i/broadcasts/1…
Română
Chjango Unchained⛓️
10.8K posts
Chjango Unchained⛓️
@chjango
madam of the robo army. investing in robotics @aexoduscapital. ex @NASA 🚀
Moonbase Katılım Ekim 2011
534 Takip Edilen13.9K Takipçiler
Chjango Unchained⛓️ retweetledi
三菱自動車がヒューマノイドロボット開発のHighlandersとフィジカルAIの量産に向けて基本合意と。
三菱の京都工場にある有休施設を活用し、2027年の量産化を目指しつつ、同社工場内での活用も検討していくとか。
大量生産して、試して、改善して…大企業とタッグでこのループの加速に期待。
日本語
U were late to crypto
U were late to AI
U missed SpaceX pre-IPO
But u know what?
Ur early to robotics.
3 irreversible forces just aligned:
1. Legacy sectors face real labor shortages now
2. Robotic "thinking" got unlocked by AI
3. Govt tailwinds committed to reindustrialization
Miles Deutscher@milesdeutscher
I think robotics is the next 10x asymmetric trade. VC investment in the sector is still 1/14th of AI, and it's about to catch up fast. Here's why I think this sector liquidity isn't slowing down anytime soon. (and how you can capitalise as an investor): The bear case on robotics for years has been the same two things: 1. AI simply hasn't been good enough 2. Hardware was too expensive to manufacture at scale Both of those constraints are breaking right now. On capability: Robotics general capabilities are rapidly improving. We are currently at the "GPT-2" moment for robotics (capable, but lacking real-world field deployment). And we're finally starting to get the first glimpse of that gap closing. @Figure_robot recently worked for 160+ consecutive hours. @weaverobotics just launched its Issac 1 humanoid bot that can handle daily tasks exceptionally well. There are many such practical examples of the dramatic improvements in robotics over the past year - this is no secret. On cost: Humanoid robot manufacturing prices have dropped from $1M+ in 2020 to $30,000-$150,000 today. Average selling prices are forecast to fall another 70% by 2030. This is the same cost curve that took solar and EV batteries from niche to mainstream in under a decade. The perfect storm is brewing right now: Robotics capabilities are growing exponentially while the cost curve simultaneously rapidly decreases. Software AI already had its moment, and I think if you're a smart investor, you'll look at physical AI. How to get exposure (nfa): → ETFs (lowest risk): $BOTZ, $ROBO, $ARKQ give you diversified sector coverage without picking individual winners. → Large caps (moderate risk): $TSLA for the Optimus bet, $AMZN for the most underrated robotics play in big tech. → Pure plays (higher risk): $OUST for robot perception/lidar, $SYM for warehouse automation. → High risk betas: $BOT (RoboStrategy) for access to private robotics companies nobody is looking at yet. There are also other interesting ways to get exposure through sectors like crypto. My full robotics article drops soon, covering every layer of how I'm personally building exposure to this sector. Be sure to follow me so you don't miss it in a few days.
English
@Player1Taco @AexodusCapital From anarcho-capitalist to robot trafficker. You can't always predict what the chrysalis will unveil.
English
@chjango @AexodusCapital This is an interesting/unexpected pivot/growth from you....
English
Chjango Unchained⛓️ retweetledi
Chjango Unchained⛓️ retweetledi
Chjango Unchained⛓️ retweetledi
I'm not missing the robotics bull run:
Perception and Sensing
$AEVA - Aeva, FMCW LiDAR
$ARBE - Arbe Robotics, 4D imaging radar
$CGNX - Cognex, machine vision
$OUST - Ouster, LiDAR
$VPG - force/load/strain sensing
$ZBRA - machine scan/vision/warehouse
Compute, Chips, and Memory (Brain)
$AMBA - edge AI vision SoCs
$AMBQ - ultra-low-power edge AI silicon
$ATOM - semi materials and IP licensing
$INDI - ADAS and automotive semis
$LSCC - low-power FPGAs
Motion, Actuation, and Precision (Body)
$ALNT - control, drives, and motors
$KLIC - semi assembly and automation
$NOVT - encoders and motion
Autonomous Mobility
$AUR - autonomous trucking
$MBLY - ADAS and autonomous driving
$TSLA - robotaxi plus Optimus humanoid
$XPEV - IRON humanoid and eVTOL
Deployment (Field/Service/Warehouse)
$KITT - subsea/ocean robotics
$RR - service robots and Dex humanoid
$SERV - sidewalk delivery
$SYM - warehouse automation systems
Surgical Robotics
$ISRG - da Vinci systems
Thematic Fund Exposure
$BOT - private names (Figure, Apptronik)
$ROBO - holdings across value chain
Let me know if I need to add anything!
Will be picking my favorite names soon.
Sun Liao@sunxliao
The humanoid robotics theme is an emerging trade I see right now and almost nobody is positioned for it correctly IMO. 🤖📈 Pay attention... most people only really know $TSLA. Optimus is real and Tesla is the demand creator that legitimizes the entire sector, but the second order trade is interesting too. Here's the chain... simplified drastically... Every humanoid robot needs eyes. Lidar and vision is the layer that lets a robot actually understand the world it is walking through. $OUST is the cleanest public lidar pure play, $MBLY is the vision and ADAS leader pivoting hard into robotics, $AMBA is the edge AI vision chip that processes everything in real time, and $AEVA, $ARBE, $CGNX round out the perception layer. Every humanoid needs precision motion. Harmonic drives, actuators, and motion control are the unsexy compounders most retail will skip right over. $VPG is the precision sensor and load cell pure play, $NOVT is motion control built specifically for robotics, $ALNT is precision motion components, and $RR is the optical sensor name that quietly shows up everywhere. Every humanoid needs a brain. The compute that runs on board has to be cheap, low power, and reliable. $LSCC is the low power FPGA that ends up inside countless edge devices, $INDI is the automotive and robotics semi nobody has on their radar yet, $AMBQ is the analog compute play, and $MRAM is the next gen memory built for exactly these workloads. Every humanoid needs a logistics use case. The first commercial deployments are not going to be in homes, they are going to be in warehouses. $SYM is the automated warehouse pure play, $ZBRA owns enterprise scanning and tracking, $SERV is sidewalk delivery robotics that doubles as data collection, and $KITT is the autonomous platform play. Every humanoid needs an industrial pedigree. The companies that already build robots for factories will be the ones supplying components and software to the humanoid OEMs. $ISRG is the surgical robotics gold standard, $KLIC is the precision assembly tooling, $HG is the heavy machinery name pivoting into robotics, and $BOT is the basket ETF if you want broad exposure in one click. And on the speculative high beta end... $ATOM is robotic software with real adoption, $XPEV has its own humanoid program coming, $AUR is autonomous trucking which is the same playbook applied to the road, and $NEO is the small cap optionality play. Pick the chokepoints. Own the picks and shovels... then wait. Will share more ideas to followers soon. NFA.
English
Chjango Unchained⛓️ retweetledi
autonomous robot driving through the field at night. no chemicals. no pesticides. just UV light killing pathogens and pests while everyone sleeps. this is @tricrobotics.
this is what chemical-free pest control looks like at scale.
English
Chjango Unchained⛓️ retweetledi
The showing we got for Isaac 1 yesterday was everything we could've hoped for
Roboticists, engineers, designers, students, kids, grandparents all in one place
Seeing how home robots will change the way we live for the better
Weave Robotics@weaverobotics
Isaac 1 Today, 4.30pm - 6.30pm 285 7th St, San Francisco
English
Within this thesis sits two portfolio companies Aexodus has invested in: @rdvrobotics and @Vaxonspace, critical infrastructure that will become the foundation for the space economy of tomorrow.
English
Getting in at the ground level of NewSpace-adjacent entrants as humanity pushes forward into the space frontier spurs a whole new economy of space infrastructure plays.
English
At the heels of the #SpaceX IPO, we offer our thesis into what likely will come next. You might be wondering, "Should I buy #SPCX?"
While the market currently looks bullish, we offer a word of caution. Boom cycles are frequently followed by bust cycles.
Aexodus@AexodusCapital
English
Chjango Unchained⛓️ retweetledi
Chjango Unchained⛓️ retweetledi
Earlier this year Yann LeCun left Meta because Mark Zuckerberg wouldn't bet the company on JEPA. Last week his group dropped the first JEPA that actually trains end-to-end from raw pixels. 15 million parameters. Single GPU. A few hours.
The timing is not a coincidence.
For four years Meta has been the house that JEPA built. LeCun published the original paper from FAIR in 2022. I-JEPA and V-JEPA came out of his lab. The architecture was supposed to be the escape hatch from LLMs, the path to robots that actually learn physics instead of hallucinating about it. Every version shipped fragile. Stop-gradients. Exponential moving averages. Frozen pretrained encoders. Six or seven loss terms that had to be hand-tuned or the model collapsed into garbage representations.
Meta kept funding LLMs. Llama shipped. Llama scaled. Llama got beat by Qwen and DeepSeek. Zuck spent $14 billion to buy ScaleAI and install Alexandr Wang. The FAIR robotics group was dissolved. LeCun's research kept winning papers and losing the product roadmap.
He left, started AMI Labs, and said publicly that LLMs were a dead end.
Now the paper. LeWorldModel. One regularizer replaces the entire pile of heuristics. Project the latent embeddings onto random directions, run a normality test, penalize deviation from Gaussian. The model cannot collapse because collapsed embeddings fail the test by construction. Hyperparameter search went from O(n^6) polynomial to O(log n) logarithmic. Six tunable knobs became one.
The downstream numbers are what should scare the robotics capex class. 200 times fewer tokens per observation than DINO-WM. Planning time drops from 47 seconds to 0.98 seconds per cycle. 48x faster at matching or beating foundation-model performance on Push-T and 3D cube control. The latent space probes cleanly for agent position, block velocity, end-effector pose. It correctly flags physically impossible events as surprising. It learned physics without being told physics existed.
Figure AI is valued at $39 billion. Tesla Optimus is mass-producing. World Labs raised $230 million to sell generative world models. Everyone in humanoid robotics is burning capital on foundation-model pipelines that plan in 47 seconds per cycle.
LeCun's group just showed you can do it with 15 million parameters on a single GPU in a few hours.
This is the Xerox PARC pattern running again. Meta had the next architecture. Meta had the scientist. Meta dissolved the robotics team, passed on the productization, and watched the exit. Three months later the lab that was supposed to be Meta's publishes the result that resets the robotics cost structure.
The paper is worth more than Alexandr Wang.
English
Chjango Unchained⛓️ retweetledi
π0.7 just proved that more robotics data can make your model worse.
what this means for robotics teams:
1. scale without metadata is pointless. without annotation density and quality scores, more data averages together conflicting strategies. the model degrades.
2. the data volume race is a trap. the teams that win won't have the most data. they'll have data their models can actually learn from.
3. PI and Standard Bots disagree on where the data should come from. they agree on what matters: structure over volume
4. the bottleneck is the loop on how fast you can organise, annotate, and feed data back into training.
Shreyas Gite@shreyasgite
English
Chjango Unchained⛓️ retweetledi
Our newest model, π0.7, has some interesting emergent capabilities: it can control a new robot to fold shirts for which we had no shirt folding data, figure out how to use an appliance with language-based coaching, and perform a wide range of dexterous tasks all in one model!
English
Chjango Unchained⛓️ retweetledi
Introducing GEN-1.
Our latest milestone in scaling robot learning.
We believe it to be the first general-purpose AI model to master simple physical tasks.
99% success rates, 3x faster speeds, adapts in real time to unexpected scenarios, w/ only 1 hour of robot data.
More🧵👇
English
Chjango Unchained⛓️ retweetledi
this weekend i learned something extremely important about writing software for manufacturing
the person writing the code must be deeply entrenched in the factory operations and nuances of the workflows
also i really dont know why you would buy any off the shelf manufacturing software anymore when you can custom build your own with all the nuance and with your own data and train your own models
English
Chjango Unchained⛓️ retweetledi
🌐 Bridging Bitchat + MeshCore: Resilient communication when infrastructure fails
Bitchat = Bluetooth mesh on phones you already have (~100m range)
MeshCore = LoRa long-range mesh (km+ with cheap hardware)
The bridge connects them. Your phone talks to the city-wide mesh network.
Perfect for disasters, protests, internet shutdowns.
Code: github.com/jooray/MeshCor…
Releases: github.com/jooray/MeshCor…
Read more: juraj.bednar.io/en/blog-en/202…
English