Tech and AIPeople are using Super Mario to benchmark AI now

People are using Super Mario to benchmark AI now

-


Thought Pokémon was a tough benchmark for AI? One group of researchers argues that Super Mario Bros. is even tougher.

Hao AI Lab, a research org at the University of California San Diego, on Friday threw AI into live Super Mario Bros. games. Anthropic’s Claude 3.7 performed the best, followed by Claude 3.5. Google’s Gemini 1.5 Pro and OpenAI’s GPT-4o struggled.

It wasn’t quite the same version of Super Mario Bros. as the original 1985 release, to be clear. The game ran in an emulator and integrated with a framework, GamingAgent, to give the AIs control over Mario.

Super Mario Bros. AI benchmark
Image Credits:Hao Lab

GamingAgent, which Hao developed in-house, fed the AI basic instructions, like, “If an obstacle or enemy is near, move/jump left to dodge” and in-game screenshots. The AI then generated inputs in the form of Python code to control Mario.

Still, Hao says that the game forced each model to “learn” to plan complex maneuvers and develop gameplay strategies. Interestingly, the lab found that reasoning models like OpenAI’s o1, which “think” through problems step by step to arrive at solutions, performed worse than “non-reasoning” models, despite being generally stronger on most benchmarks.

One of the main reasons reasoning models have trouble playing real-time games like this is that they take a while — seconds, usually — to decide on actions, according to the researchers. In Super Mario Bros., timing is everything. A second can mean the difference between a jump safely cleared and a plummet to your death.

Games have been used to benchmark AI for decades. But some experts have questioned the wisdom of drawing connections between AI’s gaming skills and technological advancement. Unlike the real world, games tend to be abstract and relatively simple, and they provide a theoretically infinite amount of data to train AI.

The recent flashy gaming benchmarks point to what Andrej Karpathy, a research scientist and founding member at OpenAI, called an “evaluation crisis.”

“I don’t really know what [AI] metrics to look at right now,” he wrote in a post on X. “TLDR my reaction is I don’t really know how good these models are right now.”

At least we can watch AI play Mario.



Source link

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Latest news

Recession In the US Confirmed? Fed Warning Triggers Bitcoin Panic

Is a recession in the US confirmed? With the 10-year bond yield once again spiking above 4.40% and...

Atua AI Strengthens XRP Cryptocurrency Backbone to Advance Financial Automation

Deeper XRP Integration Unlocks Real-Time Transactions, Smarter Asset Management, and Scalable Web3 Finance April 22, 2025 1:00 AM EDT |...

Crypto News You May Have Missed This Week | April 21, 2025

From Sony’s new web3 project to new updates from Bybit, here’s a roundup of the top crypto stories...

Advertisement

Metaplanet Makes 3rd Bitcoin Purchase in April, Expands Holdings to Over $423 Million

Japan’s first and only publicly listed Bitcoin Treasury Company, Metaplanet, has announced the acquisition of an additional 330...

Bluesky Is Rolling Out Official Verification

Starting today, Bluesky is rolling out a new verification system, complete with the familiar blue check marks popularized...

Must read

Recession In the US Confirmed? Fed Warning Triggers Bitcoin Panic

Is a recession in the US confirmed? With...

Atua AI Strengthens XRP Cryptocurrency Backbone to Advance Financial Automation

Deeper XRP Integration Unlocks Real-Time Transactions, Smarter Asset Management,...

You might also likeRELATED
Recommended to you