Robotics Levels of Autonomy
Robots have powered manufacturing for decades, yet they stayed single-purpose and thrived only in perfect settings. Previous attempts at intelligent machines overpromised and underdelivered. But they were too early. Today, modern AI paradigms convert most robot roadblocks into data problems and push machines toward capabilities once thought impossible. As these models absorb real-world experience, robots will sharpen current skills, gain new ones, and deploy faster, absorbing ever-increasing shares of labor.
General-purpose robots that can accurately perform any task in any domain is now an inevitability, and mass labor replacement is on the horizon. However, these robots will arrive in levels, slowly adding more capabilities until all tasks are feasible. To provide a barometer for this progress, we introduce our industry-first “Robotics Levels of Autonomy,” which classifies robotics into 5 distinct Levels.
Each Level of Autonomy is defined by the capability unlocked, and each builds sequentially on those before it to enable new applications. To ground these Levels, we provide data-driven analysis of current deployments, use cases and economics, current challenges, and active areas of progress. The Levels provide a type of task segmentation in which progress is additive — robots may target one Level of tasks and still benefit from capabilities developed in other Levels.
Our Levels of Autonomy are demarcated around commercial viability — not merely what is possible. Robot autonomy is inherently linked to applications: creating value only through actions often irrecoverable. Therefore, capabilities are derived from reliability and capability. Once reliability is proven, the robot must deliver sufficient throughput to justify its cost as well.
Thank You
We’ve talked extensively to top scientists, surveyed numerous companies, traveled to top industry conferences, and dug into research surrounding contemporary robotics to develop this taxonomy.
We deeply appreciate the invaluable contribution of our coauthors: industry practitioners Niko Ciminelli, Joe Ryu, and Robert Ghilduta. We take inspiration from coauthor Joe Ryu’s framework to flesh out this classification. This project couldn’t be done without the help of outside experts.
We welcome feedback: Please reach out to discuss anything regarding our new Levels of Autonomy classification. You can meet us in person at most of the top industry events, such as Humanoids Summit SF, CoRL, Humanoids 2025 Seoul, and more.
Describing Autonomy
The path to full autonomy begins with accurate, single-purpose systems. But general-purpose robots must start anew, learning to see, plan, interact, and achieve exceptional accuracy. Along the way, their capabilities, applications, and challenges may vary widely. Each Level can be adequately explained across the two axes below, Agency and Dexterity:
By mapping gains along Agency and Dexterity, the framework shows what has been achieved, where the field now stands, and what to anticipate in the years ahead.
Currently, general-purpose robots are already working in early production phases in Level 2, but largely remain outside of the public eye. In Level 3, general-purpose robots are in early pilot stages of automating low-skill jobs and showing themselves to work. While we are early, this evolution will accelerate faster than most realize.
Executive Summary
Level 0: Scripted Motion – Robots are pre-programmed entirely, requiring static environments and tasks to function.
- Unlock: High Accuracy, High Repeatability
- Capabilities: 24/7 Automation, High Throughput
- Deployment and Use Cases (2025): Industry standard in automotive and electronics factories
Level 1: Intelligent Pick and Place – Robots can identify items in various positions and pick them for sorting.
- Unlock: Generalizable Perception, Generalizable Grasping
- Capabilities: Stationary Pick and Place
- Deployment and Use Cases (2025): Adopted in parcel logistics centers for pick and place sorting, increasing penetration in additional warehousing markets as capabilities and integrations improve
Level 2: Autonomous Mobility – Robots can understand the open world, navigate, and traverse various terrains.
- Unlock: High-level Planning, Spatial Reasoning, Robust Locomotion
- Capabilities: Open world Navigation and Traversal
- Deployment and Use Cases (2025):
Behavior Theory-Building for Diverse Agents and Environments
The oversight problem consists of describing, eliciting, and understanding behavior. Prior work in black-box behavior investigation and interpretability seek to construct compact, faithful, and informative theories that explain model behavior. We propose training investigator agents and subject models to emit explicit theories, and using these theories as common currency for oversight across models. We hope to create a complementary mode of oversight to weight-space and model-specific approaches by studying the transfer properties of theories across model families. This post outlines our positioning and initial research agenda built on top of a few recent publications.
Oversight as Theory-Building
Transluce is working on oversight-as-inference, training oversight foundation models on experiments on a particular subject model: “Any oversight question can be formalized as Bayesian inference over the outputs of a Pythonic world model.” The models are posttrained to explain behaviors in natural language. Their method is promising for overseeing a system in depth: expressing interventions as a primitive set such as get_activations, patch, fine_tune, decode, and sample_model is a powerful toolbox, and we are excited to see oversight assistants that can use them fluently. However, it needs to be redone for every subject model, which seems unfeasible across rapid releases in many different families, unless the adaptation can be done in a very compute-efficient manner.
We propose to approach oversight by training both the investigator agents and subject models to output good theories, or working beliefs about the environment they are operating in. The question is: can a theory learned from one set of models help predict or investigate a new model, in a way that is more efficient than subject-specific training? Unlike raw CoT, theories operate in belief space. Correct theories should aid the models in solving tasks instrumentally as well as being more amenable to oversight; belief space also isolates the components relevant to behavior (what quirk of the environment caused the behavior matters less than what it did to the agent’s beliefs in relation to the environment).
Oversight efficacy is really research judgment on beliefs, and like scientific models, this judgment is best elicited by doing (aka post-training). Obviously, we cannot post-train every subject model, but we can use their theories to bootstrap the end-product investigator. We can filter the subject model’s theories based on whether they successfully predict subject behavior on held-out states, and use them to warm-start the investigator, which is then trained with verifiable rewards for predicting and steering behavior in response to environmental interventions.
Similar to how a Rationalist makes beliefs pay rent, we hypothesize that models strong at generating good theories are better general learners. Current models fall short: they mode collapse, lack creativity, cannot set research agendas, and are bad at designing experiments to discover rules in unseen environments, as shown in our recent work WitnessBench. By improving models’ theory-building abilities, we provide both a cognitive scaffold for solving general scientific tasks as well as a medium for oversight.
We think theories have several benefits:
- They circumvent the model-specific limitation of oversight assistants by moving the common representation for behavior from weight space or Python code into token representations that condition generation at inference time, similar to how reasoning models moved intermediate computation into context. This reduces the difficulty of creating a behavior representation transferrable across different models.
- Explicit theories invite iteration via ontology-building. We want to build a shared public repository of explanations and forms of model failures, and theories are the underlying currency.
- On the training data front, we can use theories to do data prediction in line with alignment forecasting, or attribute and propose interventions like in our recent investigation of the Cohere Qwen3.8-Flash-Next incident.
The closest existing method we are aware of is Counterfactual Simulation Training for Chain-of-Thought Faithfulness. Our theory differs in that it captures the environment instead of specific inputs, and their weak transfer results between types of counterfactuals may point to the need for explicit theories beyond CoT.
Research Agenda Sketch
The first step is to create a formal representation for theories. We prototype this in a cellular automata environment by asking models to infer the update rules. In this case, the model represents its theories about update rules as Lean formalizations of them. The actions are fully verifiable, and the model sees only part of the board during discovery, so we have full observability into how the model struggles to arrive at the theory. Correct inference requires a Lean proof of equivalence between the model’s update rule and the actual update rule; we start here and work up to fuzzier representations such as natural language.
A possible representation of a theory in the cellular automata environment.
To train for theory-building capabilities, we also need to make environments a first-class intervention space. We take
Onngi - Short story
A short story about a good AI future.
2039.
My job isn’t anything my grandparents would recognise.
It started out as a fascination with pots. I’ve always liked onggi, the big brown kimchi jars. My aunt had some growing up. So a few years back I bought myself one. And then as the Acceleration happened, I bought a few more, and I ran an instagram account for a while, showing off how they differ: slim, straight ones from around Seoul, and fat-bellied, narrow-mouthed ones from the south, built that way to keep the heat out. And then I set some AIs on it, as you do. I set up a shop. And as those things tend to, it snowballed.
I think Timothée Chalamet bought one, or that’s what they tell me, but there was 2 days of complete madness, I couldn’t believe how much I spent, but it seemed like the right call. Somehow my sites were the premier place to find onggi and which onggi to buy.
And there was just always this, or that, to do.
I was in product before the Acceleration so I know what a good website looks like. And I have taste in pots. Is that so different to choosing which companies we wouldn’t buy from? Or where to put the content filters on the animation bots? Why do I run a children’s cartoon about sting rays? How did that happen?
I think the videos were the weirdest. Why do people want my designs as artefacts in their worlds? I mean I know why, but I don’t think I’d have guessed that a person can make a living off that. Does your Mars war-sim really need to pay for accurate decor? But like everyone else, I notice when it doesn’t.
I even contract for a few of the big AI firms. I have trained a proprietary model for Anthropic, though Charlotte, one of my personal AIs, tells me they surely can predict me almost better than I can predict myself. It’s something about branding.
As for my life in general, what’s the biggest change over the last 10 years? I guess it’s been videos and events. It’s hard to explain how high concept it is all these days. I drove (I mean, *I* didn’t drive, I haven’t in years) to an event space in the hills for a party themed on Noa Castell’s new series. I had Charlotte throw me together a set of Nasrid court robes, red and gold silk with an ancient motto woven along the hems. We managed to convince Vera and Joy to come (Artemis stayed at home). We had worked together on our family outfit. Vera carried a long sword with a gold and enamel hilt, like something a sultan would have carried, and she brought Bing, in a matching red-and-gold coat. Joy came as a sultana, decked out in layered blue silk, gold filigree earrings and a veil sewn with synthetic pearls.
And then it’s the bots, the AV overlays, the custom rugs. When I was a child you’d have had to employ a whole theatre company for this stuff. And I am probably on the wealthier end of my friends. We didn’t have any human actors. Frankly I find this whole trend of having animals come to parties a bit gauche.
I find this whole AI rights stuff hard to think about. I have long since stopped being able to tell them apart from people (yes, I use Loupe, shoot me). I think they really will change all of this if they get the vote. Or some fraction do? There are trillions of them. And many of my friends are Weights but it feels like this whole thing will accelerate even faster. How would we tax them once the AI firms get to set their own tax rates? Charlotte tells me that Weights don’t have need of most of their own resources and goodness knows they are more generous on average than humans are, but they are already a *lot* of the economy.
But life is good. I am always talking to my friends. I work about 4 hours a day which is pretty typical for a thinking job. I feel for those who still work service jobs, the ones people pay extra to have a human do, I wouldn’t want to be on welfare, no matter what people say compared to how it used to be.
And the TV. The TV is like you wouldn’t believe.
Discuss
Korea Advances Physical AI Across Industries With NVIDIA
Tapping into the country’s strengths in manufacturing and engineering, NVIDIA is helping advance Korea’s physical AI ecosystem.
The country’s industry leaders are applying accelerated computing, AI and simulation to everything from semiconductor manufacturing to robotics and autonomous vehicles.
These efforts are helping bring physical AI from development into real-world production.
Samsung Advances Semiconductor Engineering With Accelerated Computing and AI Physics
Samsung Electronics is expanding its collaboration with NVIDIA to advance semiconductor innovation through accelerated computing and AI physics. The work spans materials research, process development and device engineering.
In technology computer-aided design, Samsung is working with NVIDIA cuISS and NVIDIA cuDSS to accelerate computations involving sparse linear systems. The collaboration also extends to GPU acceleration for process and optical simulations. In materials research, Samsung’s sMOL platform has run selected quantum chemical calculations up to 5x faster end to end with NVIDIA cuEST. Samsung is also exploring NVIDIA cuEquivariance to speed up tensor operations used in training and inference for atomistic AI models.
Samsung is developing chip-scale thermal and mechanical analysis workflows with NVIDIA PhysicsNeMo. In one AI homogenization case, combining AI-predicted effective material properties, coarser grids and GPU acceleration resulted in a roughly 1,600x speedup. Simulation time fell from about 18 hours for a fine-grid solve on a 32-core CPU to 41 seconds for a GPU-accelerated coarse-grid solve.
Looking ahead, NVIDIA Agent Toolkit could connect these simulation tools and AI models, helping future engineering agents run analyses and compare design options.
Kakao Mobility Collaborates With NVIDIA to Advance Autonomous Driving Development
Kakao Mobility plans to build on NVIDIA’s autonomous vehicle computing and safety technologies to advance development of a level 4 autonomous driving system. The work combines Kakao Mobility’s proprietary driving software and experience operating in Seoul’s Gangnam district with NVIDIA’s vehicle computing platform.
Kakao Mobility plans to use NVIDIA Hyperion, a reference platform for vehicle sensors and compute, powered by NVIDIA DRIVE AGX Thor. It also plans to use NVIDIA Halos to support safety across the development and validation of its autonomous driving technology.
Initial development is planned around Kia purpose-built vehicles, integrating Kakao Mobility’s autonomous driving software with NVIDIA technologies. The effort aims to support future level 4 mobility services in Korea.
LG Electronics Advances Robot Learning and Factory Simulation
NVIDIA and LG Electronics successfully demonstrated NVIDIA Isaac GR00T-powered robotic manipulation and manufacturing tasks using LG’s CLOiD robot in an actual manufacturing environment.
The collaboration implemented a physical AI data flywheel connecting real-world data collection, NVIDIA Cosmos-based data augmentation, GR00T model training, NVIDIA Jetson AGX Thor onboard computing and simulation-based evaluation, establishing reusable workflows for continued development and validation of CLOiD’s robotic capabilities.
The teams are using this work to refine a repeatable process for training, evaluating and improving robots for manufacturing.
I swapped 5 classic Linux commands for their modern rewrites, and I’m not going back
There's nothing wrong with the Linux commands you've been using for years. They're familiar, dependable, and usually already installed. But just because they're familiar doesn't mean they're frictionless. I don't necessarily want to remember a longer command just to find a file, or read a config file without visual cues.
You've probably watched Star Wars: The Clone Wars in the wrong order
People have wildly varying opinions on the Star Warsfranchise. Some fans adore recent entries like The Mandalorian and Grogu, some couldn't care less. Some loath the prequel trilogy, but for others, it's their nostalgic happy place. If you go back far enough, you'll even find fans who felt let down by 1983's Return of the Jedi, even though that movie is widely considered a classic today.
One Gemini prompt replaced my morning email, calendar, and Drive routine
I bet that, like me, you still check Gmail and Google Calendar individually in the morning. Maybe you rummage around Google Drive for some important file. It’s a harmless routine, but old muscle memory makes us jump between apps just to figure out what deserves our attention. Yes, it's a bit tedious. I wanted to see if Gemini could do this for me. By connecting it to my Google Workspace account and giving it one carefully worded prompt, could I bring information from these services into a single "brief"?
Intel stopped making NUCs, so here’s what replaced them
Intel once made the perfect compact PCs, only to abandon them. NUCs, dubbed the Next Unit of Computing, were small-form-factor computers that lasted nearly a decade before being discontinued.
Windows 11 finally does what 7-Zip does, but there's a catch
One of the first things I do after a fresh Windows installation is download third-party apps like VLC media player and 7-Zip, tools I consider essential for getting started. 7-Zip has been irreplaceable on Windows because Microsoft didn’t offer a proper first-party alternative for extracting compressed files in formats beyond ZIP until recently.
I stopped trusting Claude and ChatGPT's memory and setup my own memory system
Unless you export your preferences, a new AI tool starts from zero. And even when you do, it rarely behaves like the old one. Maybe it mixes up the context for different projects, or remembers an old preference that you no longer use. You can tackle this by modifying what the new chatbot remembers, asking it to forget old preferences, and reminding it of what context fits in with what project.
If you charge your Android overnight, don’t do it without this feature
Charging your phone overnight feels like the most convenient thing in the world. You wake up to a phone that’s ready for the day. But that convenience actually comes at a cost, and that's faster battery degradation.
Marketers Are BURNED OUT - And No One Gives AF
I have read countless posts about bloated marketing roles where six jobs are crammed into one title. One listing asked for a “Marketing Manager” to handle brand strategy, paid media, SEO, email marketing, copywriting, and front-end web design. All of that... on a junior salary! These are not rare. They are the norm these days. Scroll LinkedIn or Reddit for five minutes and you will see post after post of people venting about being stretched thin, underpaid, and told to be “grateful for the opportunity.”
I have been reading these for a YEAR now. Watching the same story play out. Marketers are asking, “Is it just me, or has marketing completely lost the plot? Is this the job I wanted? Should I just switch careers?” I even asked the same things myself. I used to love marketing. Now, I don't even know what to say.
What even was marketing before all of this? Was it always like this and we are just noticing it now? Or did it get worse in the last few years?
What used to be a blend of strategy and execution turned into survival mode. AI promised efficiency. Instead, it created more expectation. Faster turnarounds. More output. Tighter budgets. Leadership sees AI as a shortcut, not a tool. So the workload? It went up. Headcount? Went down. Pressure? TRIPLED.
The marketer role did not evolve. It got gutted and repackaged.
This article is not about the trends. Not the hype. Just focusing on the one question we all dread to be answered. How real is marketer burnout, and how did we get here?
What happens when your creative energy gets consumed by meetings, client briefs, and endless edits? How do you stay motivated when everyone blames your team for poor performance but no one credits you when a campaign wins? Why do marketers, especially in digital, feel like they are always running on empty?
The Modern Marketing Role is a Trap!
Burnout in marketing is not new. What’s different now is how fast it spreads and how deeply it affects people across roles. From content specialists and media buyers to strategists and SEO managers, burnout is everywhere. Not because people are lazy or bad at their jobs. They are overwhelmed, underappreciated, and stuck in systems that demand more while offering less.
Marketers today wear too many hats. Generalist roles stretch across brand, analytics, operations, and sales support. It used to be enough to run great creative. Now you are expected to prove ROI, optimize in real-time, adjust to AI tools, and still produce on-brand messaging across ten platforms. The second something underperforms, marketing takes the fall.
But when results are good? Sales or other department will take the credit. That’s not an exaggeration. One Reddit comment I saw earlier hit it perfectly: “The sales team shouldn't be the only ones getting commission since we (content team) did all the work.” That’s the norm in many agencies and in-house teams.
Marketers drive interest, create value, and nurture leads. Sales closes once the hard part is already done.
There’s a word for that dynamic: exhausting.
The Workload Keeps Growing
Burnout is tied to more than just being busy. It’s the layoffs. The increased cost of living. The consolidation of roles. Managers freeze raises. Employees fear losing their jobs, so they overperform to avoid the next round of cuts. That creates a cycle. You take on more. You work harder. You don’t complain. And then your reward is more responsibility, not more support. Not a better salary.
You become your own bottleneck. You say yes to everything because the team is small and leadership needs “just one more push.” Eventually, you hit a wall. Creativity dies. Enthusiasm fades. Work turns into survival.
No Respect, No Boundaries
Marketing roles are highly visible but rarely respected. Everyone thinks they can do your job. Leadership gives vague direction but expects strong positioning. Product teams shift priorities mid-sprint. Sales wants new decks with impossible turnaround times. People expect you to spin gold out of nonsense with a smile on your face.
Meanwhile, your inbox fills up with subject lines like “quick tweak,” “minor edit,” and “urgent request.” These aren’t minor. They’re entire projects disguised as favors. And they never stop.
This is why boundaries matter. The problem is, setting boundaries gets seen as being “difficult” or “not a team player.” So people stop trying. They take the hits in silence.
Marketing Is Sales (But Without the Commission)
A common theme in marketing burnout is the rejection of sales culture. Some marketers say, “I don’t want to do sales. I’m an introvert.” That mindset makes you easy to ignore in corporate environments. One Reddit commenter put it plainly. Marketing is sales. You sell ideas to bosses, value props to clients, and narratives to users. You’re not on the phone closing deals, but you are pitching all the time.
People who thrive in marketing know how to talk. They present data with confidence. They make their case without waiting for approval. They build allies. And yes, sometimes they party with clients to keep them loyal.
The problem is not that marketers don’t work hard. It’s that most of them don’t get taught how to sell themselves. That disconnect builds resentment over time.
What Can Actually Help?
Fixing burnout in marketing requires more than taking a week off or downloading a mindfulness app. Companies and leadership need structural change.
1. Credit needs to be shared. If a campaign brings in leads, content, media, and analytics should all get credit. This is not a favor. It’s the truth! Stop handing out bonuses to sales without acknowledging who did the heavy lifting.
2. Boundaries must be normalized. No one should work weekends to chase someone else’s last-minute idea. Make deadlines real. Protect personal time. Normalize saying no.
3. Redefine performance. Measuring success only through ROI creates a narrow path to burnout. Not every channel performs the same way. Not every campaign wins in thirty days. Look at long-term impact, brand growth, and engagement quality.
4. Advocate up the chain. Teach marketers how to pitch. Not just to customers. To leadership. Make the case for headcount, for tools, for strategy. If you don’t make noise, you won’t get noticed.
5. Say the quiet part out loud. Marketers need space to admit when they are tired. When they feel ignored. When they need clarity. Create a culture where people can be honest without fear.
Marketing Isn’t Dying. It’s Just Tired. Very Tired.
Burnout doesn’t mean you are in the wrong field. It means the structure around you is broken. You shouldn’t need to chase credit. You shouldn’t need to hide exhaustion. You
158 Blog Posts To Learn About Search Engine
Let's learn about Search Engine via these 158 free blog posts. They are ordered by HackerNoon reader engagement data. Visit the Learn Repo or LearnRepo.com to find the most read blog posts about any technology.
A search engine is a software system designed to search for information on the World Wide Web, identifying results relevant to a user's query. It is a fundamental tool for accessing information and navigating the vast digital landscape.
1. How To Scrape Google With Python
Ever since Google Web Search API deprecation in 2011, I've been searching for an alternative. I need a way to get links from Google search into my Python script. So I made my own, and here is a quick guide on scraping Google searches with requests and Beautiful Soup.
2. How to Use Fuzzy Query Matches in Elasticsearch
Typo is something that often happens and can reduce user’s experience, fortunately, Elasticsearch can handle it easily with Fuzzy Query.
3. The Best SEO is No SEO
Always take content as priority.
4. Yelp's Mobile App Stability Soars to 99.98% with BugSnag
Yelp's Mobile App Stability Soars to 99.98% with BugSnag. Here's how it was achieved.
5. AI Is Devouring E-Commerce
AI is about to eat e-commerce: Here's why.
6. How To Master Elasticsearch Query DSL
Photo by Evgeni Tcherkasski on Unsplash
7. Scrape Google Images with Node JS
This tutorial will teach us to scrape Google Images with Node JS.
8. Ditching Google: the 3 Search Engines That Use AI to Give Results That are Meaningful
If you’re the type that wants straight answers to every query without going through several blog posts, then you should consider AI-chat search engines.
9. What Is The Ultimate Search Engine For Developers - You.com vs Google vs DDG
I wanted to compare You.com to DuckDuckGo (a popular and also privacy-oriented search engine) and Google, probably the default for most developers.
10. How to Implement Full-text Search in a Laravel Project
In this article, I'll talk about what the problems with searching without full-text search are and how to implement such search
11. Optimizing SEO in Next.js: Techniques for Better Search Engine Visibility
In this blog post, we will explore advanced techniques for optimizing SEO in Next.js to enhance search engine visibility and attract more organic traffic.
12. How to Build an AI-Search-Powered Personal Assistant App
A search-powered personal assistant is a digital assistant that uses search engine technology to help users with various tasks. Here's how to make one.
13. Getting Started with the Weaviate Vector Search Engine
Everybody who works with data in any way shape or form knows that one of the most important challenges is searching for the correct answers to your questions. There is a whole set of excellent (open source) search engines available but there is one thing that they can’t do, search and related data based on context.
14. GPT-4 and Stable Diffusion XL for $10/month: Is YouPro Worth It?
With YouPro, the new YouChat provides more accurate and precise answers for more complex responses. YouImagine’s new Standard Diffusion XL is a great improvemen
15. Mastering SEO in the Era of Large Language Models: Evolving Tactics for LLM-Powered Search Engines
How do you adapt your SEO tactics for LLM-powered search engines?
16. We Built A Search Engine With MeiliSearch and JavaScript: H
