What Strawberry, OpenAI’s New AI That Can “Think” Would Look Like (And What Will Happen to ChatGPT)

After accessing a series of secret documents, a media outlet has discovered the Strawberry project, a new artificial intelligence from OpenAI, creator of ChatGPT, which “thinks” before responding to humans. Its launch is said to take place in two weeks.

A Artificial intelligence who “thinks” before answering. That’s how it would be Strawberry, the supposed new AI that is going to be launched OpenAI —parent company of ChatGPT— and which will once again revolutionize this technology, with a tool within everyone’s reach.

The “rumors” about this new AI model are strong: the media The information And Reuters they published this In the next two weeks, the technology company will launch Strawberry, a chatbot that will this time have the ability to “reason” before answering our questions.

So, two years after the dizzying change that ChatGPT has caused – launched in November 2022 – in humanity and in the way we interact with this technology, Everything indicates that OpenAI plans to surprise the world with a new system that no other company has explored so far.

However, This time it won’t be free.

Additionally, some have asked, What will happen to ChatGPT?

This is everything we know about OpenAI’s Strawberry until now.

OpenAI
What Strawberry, OpenAI’s New AI That Can “Think” Would Look Like (And What Will Happen to ChatGPT)

What is Strawberry, OpenAI’s supposed new AI?

This new model of Artificial Intelligence, named Strawberry will not replace ChatGPT, but will be integrated independently.

Strawberry would add to the big improvements OpenAI has implemented with ChatGPT : as the tech media explain, Xataka The company had already announced “a promising vocal mode that recalls the future proposed by the film.” Her“.

The difference that Strawberry would make is that, compared to conventional ChatGPT, This new system could “think” before answering any question.

Nevertheless, The exact concept of how this AI will “think” remains a closely guarded secret by OpenAI.

Anyway, Reuters He had access to some internal documents of the brand where he discovered the name Strawberry, in the middle of this year 2024, and some anonymous sources linked to the company confirmed that indeed, There is this gourmet project that will have a “large capacity”.

According to these same sources, Strawberry AI will be able to autonomously “reason and navigate” the Internet to be able to develop more complete and precise answers. It was supposedly called “in-depth research” internally, they said in the same media.

Artificial intelligence
What Strawberry, OpenAI’s New AI That Can “Think” Would Look Like (And What Will Happen to ChatGPT)

That’s to say, Strawberry could be a kind of “assistant” that could help us in more complex searches, compare results and find complete answers. something we usually do manually when we search for something specific on the web.

Responses would take between 10 and 20 seconds, during which time the AI ​​would conduct a “thorough investigation.” consult current internet sources . This may not be convenient for those looking for a quicker answer, but it may be convenient for those doing more detailed searches.

According to The informationStrawberry will arrive earlier than expected: Although this information has not been made official by OpenAI, the media assured that the launch of this tool would take place “in the next two weeks.”

In addition, Xataka he added thatIt is possible that as an independent service from ChatGPT it is accessible in a ChatGPT drop-down menu.

What is it yes, that’s it The information he announced that Strawberry could have a high price tag, even higher than ChatGPT, and could be available as an “annual subscription.”

“How much would you be willing to pay for ChatGPT each month? $50? $75? How about $200 or $2,000?” the outlet asked in an article.

As his journalists would have heard, OpenAI executives “have talked about expensive subscriptions for the next big language models, such as Strawberry, focused on reasoning, and a new flagship LLM model, called Orion.

“How much customers are willing to pay for the AI ​​chatbot matters not only to OpenAI, but also to competitors offering similar products, including Google, Anthropic and others,” they wrote in the outlet.

ChatGPT
What Strawberry, OpenAI’s new AI that can “think” would look like (and what will happen to ChatGPT). Photo: REUTERS/Dado Ruvic

In fact, during early internal discussions at OpenAI, one “person with direct knowledge of the numbers” said: The information that Subscription prices of up to $2,000 per month (about 1,900,000 Chilean pesos) were reportedly considered.

Despite this, the media said they had “serious doubts that the final price would be so high.”

In addition, With Strawberry, OpenAI would continue to be at the forefront of the increasingly competitive Artificial Intelligence race. Some companies, such as Google continue to work on colossal language models, so this new system could be one of the many surprises in store for humanity in the years to come.

A few months ago, Brad Lightcap, COO at OpenAI, announced in a conversation with Financial Time that The company is already looking to create more complex AI tools that can solve “harder problems” through reasoning.

“We’re going to start seeing that AI can do more complex tasks in more sophisticated ways. “We are only just beginning to scratch the surface of the reasoning capacity of these models,” declared.

AI
What Strawberry, OpenAI’s New AI That Can “Think” Would Look Like (And What Will Happen to ChatGPT)

In this sense, according to the media Wired , Analysts expect Strawberry’s alleged reasoning capabilities to “mark the next big breakthrough in AI.”

And that’s it, For the first time, technology will be able to “make judgments, reason, communicate in natural language and learn from its environment.”

But, for other experts, This could also be a double-edged sword, as there is still a latent fear that AI could outperform humans and “reveal itself,” or that this technology could be used by malicious people.

Noah Goodman, an academic at Stanford University told the same media that this new advance “It’s exciting and terrifying at the same time. If things continue in this direction, as humans we’re going to have some serious questions to ask ourselves.”

Source: Latercera

Related articles

Comments

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Share article

Latest articles

Newsletter

Subscribe to stay updated.