The ARC Prize organization designs benchmarks which are specifically crafted to demonstrate tasks that humans complete easily, but are difficult for AIs like LLMs, “Reasoning” models, and Agentic frameworks.
ARC-AGI-3 is the first fully interactive benchmark in the ARC-AGI series. ARC-AGI-3 represents hundreds of original turn-based environments, each handcrafted by a team of human game designers. There are no instructions, no rules, and no stated goals. To succeed, an AI agent must explore each environment on its own, figure out how it works, discover what winning looks like, and carry what it learns forward across increasingly difficult levels.
Previous ARC-AGI benchmarks predicted and tracked major AI breakthroughs, from reasoning models to coding agents. ARC-AGI-3 points to what’s next: the gap between AI that can follow instructions and AI that can genuinely explore, learn, and adapt in unfamiliar situations.
You can try the tasks yourself here: https://arcprize.org/arc-agi/3
Here is the current leaderboard for ARC-AGI 3, using state of the art models
- OpenAI GPT-5.4 High - 0.3% success rate at $5.2K
- Google Gemini 3.1 Pro - 0.2% success rate at $2.2K
- Anthropic Opus 4.6 Max - 0.2% success rate at $8.9K
- xAI Grok 4.20 Reasoning - 0.0% success rate $3.8K.
(Logarithmic cost on the horizontal axis. Note that the vertical scale goes from 0% to 3% in this graph. If human scores were included, they would be at 100%, at the cost of approximately $250.)
https://arcprize.org/leaderboard
Technical report: https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf
In order for an environment to be included in ARC-AGI-3, it needs to pass the minimum “easy for humans” threshold. Each environment was attempted by 10 people. Only environments that could be fully solved by at least two human participants (independently) were considered for inclusion in the public, semi-private and fully-private sets. Many environments were solved by six or more people. As a reminder, an environment is considered solved only if the test taker was able to complete all levels, upon seeing the environment for the very first time. As such, all ARC-AGI-3 environments are verified to be 100% solvable by humans with no prior task-specific training
There’s also the concept of consciousness without memory. What’s that like? Being able to experience the current moment but having no memory of any past experiences - including your experience one second ago.
Or here’s a scary thought: what if general anesthesia doesn’t actually switch off consciousness but simply blocks new memories from forming? You could experience the full horror of being awake during surgery but remember none of it. From the perspective of “now,” that would be functionally the same as never having experienced it at all.
Then there are those extremely weird recordings from split-brain studies. Back when grand mal seizures were treated by cutting the corpus callosum - the bridge between the two brain hemispheres - to stop the “storm” from spreading. On the surface these patients seemed completely normal after the operation, but some really strange stuff shows up when you start testing them properly.
There’s a way you can communicate with each hemisphere independently without the other one knowing. The left hemisphere controls the right side of the body, the right hemisphere the left side. You can flash text on the left side of the visual field (which only the right, non-verbal hemisphere sees), then hand them a pen and let the left hand (controlled by the right hemisphere) answer questions by writing. Turns out that the two halves often don’t agree on things. Ask the right hemisphere what it wants to do for a living and you’ll get a different answer than what the left hemisphere says out loud. Or you can give the right hemisphere a task - “go get a glass of water” - and when you ask the left hemisphere why it did that, it just makes up an explanation. “I was thirsty,” it’ll say, even though the researchers know that’s not true. It genuinely seems like there are two separate consciousnesses running in the same brain at the same time. The big question is: were they there all along, or does the second one only emerge once the connection between them is cut?
And yeah, this is all stuff I’ve absorbed from podcasts covering these topics - mostly from Sam Harris. I’m just naturally really curious about the human mind, and I’m pretty experienced with meditation as well, so I probably pay attention to my day-to-day conscious experiences about 1% more than the average person. I’m however not in any way expert on this. It’s not even remotely related to what I do for living.