Claude AI of Anthropor Pokemon plays on Twitch … slowly


On Tuesday afternoon, I launched the Antarbur Claude plays Pokemon On Twitch, a direct flow of the latest AI for anthropology, Claude 3.7 SonataPlay the red Pokemon game. It has become a wonderful experience of some kind, as it displays the capabilities of artificial intelligence today and people’s reactions to it.

Artificial intelligence researchers used all kinds video gamesfrom Street fighter to pisionaryTo test new models – it is often more than entertainment of interest. But Anthropor said that Pokemon has proven to be a useful standard for Claude 3.7 Sonnet, which can effectively can “He thinks” Through the types of puzzles that the game contains.

Like Openai O3-Mini And Deepseek R1Claude 3.7 Sonnet “Thinking” can think about it through difficult challenges, such as playing a video game designed for children. While the predecessor of the unusual model, Claude 3.5 SonataFailing at the beginning of Pokemon Reed – leaving the player’s house in the town of Pallite – Claude 3.7 Sonnet managed to win three windows of the gym leaders.

Claude is struggling to get out of a building in Red Pokemon.Image credits:Claude plays Pokemon on Nashil

The latest Claude is still having a problem, though. Hours after the Nashl Movement, the model was deterred by a rocky wall, which he could not walk in regardless of how difficult it was.

One of the users of the situation in this way summarized: “Who will win, a computer was placed with thousands of hours in its programming, or one rock wall?”

In the end, Claude realized that he could move around the wall.

On the one hand, it is frustrated. SlowThinking through every step with temporary meditation. However, he is strangely convincing. The left of the stroke shows the “thinking process”, while the right shows the way to play in actual time.

At one point, Claude tried to locate Professor Ok inside his laboratory, but he felt confused, because there was another NPCS in the scene.

“I have noticed that there is a new figure that has appeared under – a black hair and what appears to be a white coat in the coordinates (2, 10),” Claude wrote. “This may be Professor Ok! Let me go down and talk to him.”

Claude then moved to talking to NPC other than the processor – the NPC that the model has caused several times before. Some thousand people have started to chat with Antsy. The others, especially those who were watching the current for more than a few minutes, were less anxious.

A person wrote in the chat “Cold Men”. “Before we got out and entered Oak Lab, like 10 times before understanding how to go ahead.”

Red Human Pokemon
Image credits:man

For Twitch users for a long time, the coordination of the human current may feel nostalgia. More than a decade ago, millions of people tried to play Pokémon Red at once in an online social experience of its kind Nashl play Pokemon. Each user can control the via Twitch chat, which leads to an expected chaotic toys.

Some artificial intelligence researchers were martyred that Pokemon manipulated an inspiration for their work. In October 2023, Peter Widin, a Sayattle -based software engineer, posted a video on YouTube, explaining in detail how he trained the algorithm learning to reinforce to play Pokemon. He spent artificial intelligence to him 50,000 hours playing the game Before you learn to move successfully. One of the challenges was that artificial intelligence preferred to admire the landscape instead of actually playing the game.

“Re -legislation” of artificial intelligence from Twitch plays Pokemon like Whinken’s and Anthropic’s entertaining, but a little sweet waters at the same time. The original flow was a pivotal moment in the history of Nashal because it gathered people in an unexpected way. Everyone was in the same team, as he is working towards the goal of getting the player’s personality to stop running in circles and actually progress during the game.

In 2025, it seems that we are no longer teammates, but spectators, we are seeing the Amnesty International model trying to play a game that disrupts many of us when our age was five years old. It is a miniature of Amnesty International’s motives for a larger direction: our online experiences are transmitted from joint and community activities to solitary activities.

Leave a Reply

Your email address will not be published. Required fields are marked *