Today we're releasing a developer preview of our next-gen benchmark, ARC-AGI-3. The goal of this preview, leading up to the full version launch in early 2026, is to collaborate with the community. We invite you to provide feedback to help us build the most robust and effective , page 182

Today we're releasing a developer preview of our next-gen benchmark, ARC-AGI-3.

The goal of this preview, leading up to the full version launch in early 2026, is to collaborate with the community. We invite you to provide feedback to help us build the most robust and effective benchmark possible.

ARC-AGI-3 is the continuation of the ARC series of AI benchmarks:

• Focus on generalization and adaptation to novelty
• Easy for humans, yet extremely difficult for AI
• Built exclusively on Core Knowledge priors, with no other domain-specific knowledge required

The biggest evolution is a complete shift to an interactive format. ARC 3 is a collection of unique, novel games set in the ARC grid world. To succeed, an AI must learn on the fly through efficient trial and error. There are no instructions, so everything is up for you to figure out:

• What are the underlying concepts and mechanics?
• What do the controls do?
• What is the goal, and how do I achieve it?

We are directly probing an AI's ability to efficiently explore, learn, and plan when faced with a completely unknown task. So far, all systems we've tested score 0. Yet all these games are fully solvable by humans in a few minutes with no prior training.

点击图片查看原图

1周内 1个月内 1年内全部时间

“The state of most social platforms right now is that users are locked in and developers are locked out. We want to build something that makes sure users have the freedom to move and developers have the freedom to build.”
— Jay Graber, CEO
时政
( bsky.app)

1年前 • bluesky 0 评论

ELON MUSK: "This is not the end of DOGE, but really the beginning ... The DOGE team will only grow stronger overtime ... I am confident over time, we will see $1 trillion dollars of waste and fraud reduction."
Thank you, Elon Musk 🫡🇺🇸
btc
( twitter.com)

8个月前 • Charlie Kirk • 下载视频 0 评论

00:02:21

I watched this like 3 times because it is so mind blowing
“This is a microprocessor, how did humans make this. How as human beings did we even come close to making something like this”
He takes a microscope and keeps zooming:
btc
( twitter.com)

8个月前 • Wall Street Apes • 下载视频 0 评论

00:00:58

Elon Musk says AI won't just lead to universal basic income -- but universal high income.
Tens of billions of humanoid robots could 10× the global economy. They won't just help. They’ll unlock abundance.
Everyone gets what they want, if we get it right.
btc
( twitter.com)

8个月前 • vitrupo • 下载视频 0 评论

00:02:10

Help the FBI protect children. As the Attorney General has made clear, we will protect our children and hold accountable those who mutilate them under the guise of gender-affirming care.
Report tips of any hospitals, clinics, or practitioners performing these surgical
时政
( twitter.com)

8个月前 • FBI 0 评论

Europe, together with America, has better weapons than Russia. We also have stronger tactical solutions – our operation “Spiderweb” yesterday proved that. Russia must feel what its losses mean. That is what will push it toward diplomacy.
And when Russia takes losses in this
时政
( twitter.com)

8个月前 • Volodymyr Zelenskyy / Володимир Зеленський • 下载视频 0 评论

00:04:29

With President Trump, we KNOW who is in CHARGE.
Under Biden, America was rudderless—drifting without leadership or accountability.
That’s why Afghanistan collapsed. That’s why war broke out in Ukraine. That’s why Iran was emboldened and Hamas launched its brutal assault on
btc
( twitter.com)

8个月前 • Wesley Hunt • 下载视频 0 评论

00:00:59

We must free ourselves from this vicious, seven-step cycle:
(1) Government expands
(2) Larger government becomes abusive
(3) Voters demand reform
(4) Leftists insist that “people will die” if reform is achieved
(5) No one questions their assertions, lest they seem
btc
( twitter.com)

8个月前 • Mike Lee 0 评论

I unapologetically stand for free speech, peaceful demonstrations, and immigration—but this is not that.
This is anarchy and true chaos.
My party loses the moral high ground when we refuse to condemn setting cars on fire, destroying buildings, and assaulting law enforcement.
btc
( twitter.com)

7个月前 • U.S. Senator John Fetterman 0 评论

I was a reporter and anchor for ABC News. And as you may have heard … I'm not there anymore.
Now I'm on Substack, that amazing space, and I can't wait to get into the important work that we all have to do.
时政
( twitter.com)

7个月前 • Terry Moran 🇺🇸 • 下载视频 0 评论

00:00:41

Solar + Storage will be the majority of meeting growing energy needs. The sun is a reliable giant fusion reactor and we just need to catch a minuscule amount of it. The largest Supercharger in the world will be powered by the sun.
btc
( twitter.com)

7个月前 • Max 0 评论

We just unveiled Grok 4, the world’s smartest artificial intelligence. 🧵
Grok 4 outperforms all other models on the ARC-AGI benchmark, scoring 15.9% - nearly double that of the next best model - and establishing itself as the most intelligent AI to date.
btc
( twitter.com)

6个月前 • xAI • 下载视频 0 评论

00:01:22

NEWS: As of 2024, Tesla drivers using Autopilot technology (with active driver supervision) achieved a safety record nearly 10x higher than the U.S. national average across all automobile makers and models.
"We expect to replicate and accelerate this level of safety with the
btc
( twitter.com)

7个月前 • Sawyer Merritt 0 评论

xAI continues to lead the way with cutting-edge technology. We have commenced construction on the world’s large ceramic membrane bioreactor (MBR) to sustainably supply 13 million gallons per day of cooling water to our Colossus Supercomputer and neighboring industrial plants in
btc
( twitter.com)

6个月前 • xAI Memphis 0 评论

New blog post about asymmetry of verification and "verifier's law":
Asymmetry of verification–the idea that some tasks are much easier to verify than to solve–is becoming an important idea as we have RL that finally works generally.
Great examples of
时政
( www.jasonwei.net)

6个月前 • Jason Wei 0 评论

This is how #DeepSeek censors my new graphic novel 'YOU MUST TAKE PART IN REVOLUTION'
It also gives some authentic Ministry of Foreign Affairs of China answers when we ask its opinions on me and my co-author
Pre-order the book here:
时政
( Badiucao.com)

6个月前 • 巴丢草 Bad ї ucao 0 评论

Things we are looking forward in 2025 with Grok that's currently known
- Native Video input
- Grok 4.20
- Grok 4 Code
- Grok Mac App
- Baby Grok
- Big Video Generation Model (~October)
- New Image Generation Model
- Code Editor (unconfirmed)
- New AI Companions
- DLCs for
btc
( twitter.com)

5个月前 • Tech Dev Notes 0 评论

Such a wonderful moment with my friends coming to opening of my Melbourne exhibition ‘Disagree Where We Must’.
My art must address darkness and crisis in China Australia and around the world.
And those support and inspiration from human rights heros like
时政
( twitter.com)

5个月前 • 巴丢草 Bad ї ucao 0 评论

🚨 We just minted $205,000,000 of USD1 for the WLFI Treasury🚨
The fastest-growing stablecoin of all time continues to scale. Backed 1:1 by USD and U.S. Treasuries. Transparent. Reliable. Built for the future of global finance.
$USD1 ☝️| $WLFI 🦅
btc
( twitter.com)

5个月前 • WLFI 0 评论

There are now 12 people around the world with Neuralink implants. Collectively, they’ve had their devices for 2,000 days and accumulated over 15,000 hours of use.
We look forward to continuing to explore the possibilities of neural interfaces with all of our participants!
btc
( twitter.com)

4个月前 • Neuralink 0 评论

The algorithm will be purely AI by November, with significant progress along the way.
We will open source the algorithm every two weeks or so.
By November or certainly December, you will be able to adjust your feed dynamically just by asking Grok.
btc
( twitter.com)

4个月前 • Elon Musk 0 评论

1⃣ We will start our mint in 30 hours
我们明晚香港时间9点开始mint
2⃣ Check the mint price
请确认mint的价格、数量
3⃣ And the mint site/m…
推特中文圈
( twitter.com)

2年前 • Dungeon Relics 0 评论

To all the #media, #VC and #Community We look forward to seeing you again.
非常感谢香港认识的每一位！
我作为Jasmy的大中华区大使会努力推广Jasmy。
币圈
( twitter.com)

2年前 • The JASMY Movement 0 评论

We Stand with Dalai Lama🙏💞
我们和达赖喇嘛尊者站在一起🙏🌈
昨天多伦多地区多个藏人团体举行
#WeStandWithDalaiLama
的游行集会
时政
( twitter.com)

2年前 • 盛雪SHENG Xue👌推翻共产党死磕习始皇 0 评论

北同文化 Beijing LGBT+ announces it is suspending operations "我们非常遗憾地告知大家，因不可抗力，北同文化于今日终止运营 We regret to inform everyon…
推特中文圈
( twitter.com)

2年前 • Bill Bishop 0 评论

Today we're releasing a developer preview of our next-gen benchmark, ARC-AGI-3. The goal of this preview, leading up to the full version launch in early 2026, is to collaborate with the community. We invite you to provide feedback to help us build the most robust and effective btc ( twitter.com )

btc

“The state of most social platforms right now is that users are locked in and developers are locked out. We want to build something that makes sure users have the freedom to move and developers have the freedom to build.” — Jay Graber, CEO 时政 ( bsky.app)

时政

ELON MUSK: "This is not the end of DOGE, but really the beginning ... The DOGE team will only grow stronger overtime ... I am confident over time, we will see $1 trillion dollars of waste and fraud reduction." Thank you, Elon Musk 🫡🇺🇸 btc ( twitter.com)

btc

I watched this like 3 times because it is so mind blowing “This is a microprocessor, how did humans make this. How as human beings did we even come close to making something like this” He takes a microscope and keeps zooming: btc ( twitter.com)

btc

Elon Musk says AI won't just lead to universal basic income -- but universal high income. Tens of billions of humanoid robots could 10× the global economy. They won't just help. They’ll unlock abundance. Everyone gets what they want, if we get it right. btc ( twitter.com)

btc

Help the FBI protect children. As the Attorney General has made clear, we will protect our children and hold accountable those who mutilate them under the guise of gender-affirming care. Report tips of any hospitals, clinics, or practitioners performing these surgical 时政 ( twitter.com)

时政

时政

btc

We must free ourselves from this vicious, seven-step cycle: (1) Government expands (2) Larger government becomes abusive (3) Voters demand reform (4) Leftists insist that “people will die” if reform is achieved (5) No one questions their assertions, lest they seem btc ( twitter.com)

btc

I unapologetically stand for free speech, peaceful demonstrations, and immigration—but this is not that. This is anarchy and true chaos. My party loses the moral high ground when we refuse to condemn setting cars on fire, destroying buildings, and assaulting law enforcement. btc ( twitter.com)

btc

I was a reporter and anchor for ABC News. And as you may have heard … I'm not there anymore. Now I'm on Substack, that amazing space, and I can't wait to get into the important work that we all have to do. 时政 ( twitter.com)

时政

Solar + Storage will be the majority of meeting growing energy needs. The sun is a reliable giant fusion reactor and we just need to catch a minuscule amount of it. The largest Supercharger in the world will be powered by the sun. btc ( twitter.com)

btc

We just unveiled Grok 4, the world’s smartest artificial intelligence. 🧵 Grok 4 outperforms all other models on the ARC-AGI benchmark, scoring 15.9% - nearly double that of the next best model - and establishing itself as the most intelligent AI to date. btc ( twitter.com)

btc

NEWS: As of 2024, Tesla drivers using Autopilot technology (with active driver supervision) achieved a safety record nearly 10x higher than the U.S. national average across all automobile makers and models. "We expect to replicate and accelerate this level of safety with the btc ( twitter.com)

btc

xAI continues to lead the way with cutting-edge technology. We have commenced construction on the world’s large ceramic membrane bioreactor (MBR) to sustainably supply 13 million gallons per day of cooling water to our Colossus Supercomputer and neighboring industrial plants in btc ( twitter.com)

btc

New blog post about asymmetry of verification and "verifier's law": Asymmetry of verification–the idea that some tasks are much easier to verify than to solve–is becoming an important idea as we have RL that finally works generally. Great examples of 时政 ( www.jasonwei.net)

时政

This is how #DeepSeek censors my new graphic novel 'YOU MUST TAKE PART IN REVOLUTION' It also gives some authentic Ministry of Foreign Affairs of China answers when we ask its opinions on me and my co-author Pre-order the book here: 时政 ( Badiucao.com)

时政

Things we are looking forward in 2025 with Grok that's currently known - Native Video input - Grok 4.20 - Grok 4 Code - Grok Mac App - Baby Grok - Big Video Generation Model (~October) - New Image Generation Model - Code Editor (unconfirmed) - New AI Companions - DLCs for btc ( twitter.com)

btc

Such a wonderful moment with my friends coming to opening of my Melbourne exhibition ‘Disagree Where We Must’. My art must address darkness and crisis in China Australia and around the world. And those support and inspiration from human rights heros like 时政 ( twitter.com)

时政

🚨 We just minted $205,000,000 of USD1 for the WLFI Treasury🚨 The fastest-growing stablecoin of all time continues to scale. Backed 1:1 by USD and U.S. Treasuries. Transparent. Reliable. Built for the future of global finance. $USD1 ☝️| $WLFI 🦅 btc ( twitter.com)

btc

There are now 12 people around the world with Neuralink implants. Collectively, they’ve had their devices for 2,000 days and accumulated over 15,000 hours of use. We look forward to continuing to explore the possibilities of neural interfaces with all of our participants! btc ( twitter.com)

btc

The algorithm will be purely AI by November, with significant progress along the way. We will open source the algorithm every two weeks or so. By November or certainly December, you will be able to adjust your feed dynamically just by asking Grok. btc ( twitter.com)

btc

1⃣ We will start our mint in 30 hours 我们明晚香港时间9点开始mint 2⃣ Check the mint price 请确认mint的价格、数量 3⃣ And the mint site/m… 推特中文圈 ( twitter.com)

推特中文圈

To all the #media, #VC and #Community We look forward to seeing you again. 非常感谢香港认识的每一位！ 我作为Jasmy的大中华区大使会努力推广Jasmy。 币圈 ( twitter.com)

币圈

We Stand with Dalai Lama🙏💞 我们和达赖喇嘛尊者站在一起🙏🌈 昨天多伦多地区多个藏人团体举行 #WeStandWithDalaiLama 的游行集会 时政 ( twitter.com)

时政

北同文化 Beijing LGBT+ announces it is suspending operations "我们非常遗憾地告知大家，因不可抗力，北同文化于今日终止运营 We regret to inform everyon… 推特中文圈 ( twitter.com)