Revolutionary AI Model Makes Strides in Tech World
A tech giant is seeing substantial advancements in the field of artificial intelligence (AI), according to a company insider. The firm's latest AI model, playfully codenamed 'Watermelon', is said to be on par with the esteemed GPT-5.5 model, a flagship AI model of a prominent competitor.
The progress was reported by the head of the company's superintelligence division during an internal meeting. The claim was reportedly based on widely recognized AI model benchmarks, although it remains unclear which exact benchmarks were referenced.
Watermelon: The Next Big Thing in AI
'Watermelon' is the successor to the company's previous model, 'Avocado'. During the meeting, it was revealed that 'Watermelon' utilizes significantly more computational power than its predecessor. 'Avocado' was the internal name given to Muse Spark, the first model in a series that was launched by the company earlier this year.
The head of the superintelligence division also hinted at the progress in a public post. He announced that an update to Muse Spark is imminent, featuring major enhancements in coding and agentic capabilities. These improvements are aimed at narrowing the gap with competing models. In response to a query about when the company would have a coding model equal to another competitor's model, 'Claude Opus', he responded optimistically, assuring that it would be "pretty soon" and promising that users will appreciate what the company is "cooking".
AI Goals and Challenges
The tech company has consistently aimed to match, if not surpass, the AI capabilities of rivals. Despite substantial investments in chips, data centers, and talent, convincing developers and customers of the superiority of its models has been an uphill battle.
If the claims about 'Watermelon' are accurate, it may be the most tangible evidence yet of the company's investments and aggressive recruitment strategy starting to yield results. This is despite the fast-paced and relentless nature of the AI race.
The GPT-5.5 model, which 'Watermelon' is said to rival, is a formidable AI model that was released earlier this year by a competitor. The competitor has since unveiled an even more powerful model, GPT-5.6, although it has not yet been widely released due to government requests.
Previous Efforts and Future Plans
Earlier this year, the tech firm unveiled the first in a series of models known as Muse Spark. While it performed well on benchmarks, it could not match or exceed the performance of models from competitors.
The company's relentless pursuit of AI supremacy is led by the firm's founder, who has been pushing heavily for the company to lead the AI race. Last year, he appointed the current head of the superintelligence division and rebranded the company's AI division as Meta Superintelligence Labs.
Within the company, the head of the division oversees a team of top-tier AI researchers, as well as other AI initiatives, including a recent hardware push. The company has reportedly offered hundreds of millions of dollars to top AI talent in an effort to recruit them.
This aggressive recruitment comes as the company escalates its spending on infrastructure. Investors were informed this year that the firm plans to spend between $125 billion and $145 billion on chips, data centers, and other infrastructure. This marks an increase from an initial forecast of $115 billion to $135 billion, due to rising component costs and additional data center investments.
A tech giant is seeing substantial advancements in the field of artificial intelligence (AI), according to a company insider. The firm's latest AI model, playfully codenamed 'Watermelon', is said to be on par with the esteemed GPT-5.5 model, a flagship AI model of a prominent competitor.
The progress was reported by the head of the company's superintelligence division during an internal meeting. The claim was reportedly based on widely recognized AI model benchmarks, although it remains unclear which exact benchmarks were referenced.
Watermelon: The Next Big Thing in AI
'Watermelon' is the successor to the company's previous model, 'Avocado'. During the meeting, it was revealed that 'Watermelon' utilizes significantly more computational power than its predecessor. 'Avocado' was the internal name given to Muse Spark, the first model in a series that was launched by the company earlier this year.
The head of the superintelligence division also hinted at the progress in a public post. He announced that an update to Muse Spark is imminent, featuring major enhancements in coding and agentic capabilities. These improvements are aimed at narrowing the gap with competing models. In response to a query about when the company would have a coding model equal to another competitor's model, 'Claude Opus', he responded optimistically, assuring that it would be "pretty soon" and promising that users will appreciate what the company is "cooking".
AI Goals and Challenges
The tech company has consistently aimed to match, if not surpass, the AI capabilities of rivals. Despite substantial investments in chips, data centers, and talent, convincing developers and customers of the superiority of its models has been an uphill battle.
If the claims about 'Watermelon' are accurate, it may be the most tangible evidence yet of the company's investments and aggressive recruitment strategy starting to yield results. This is despite the fast-paced and relentless nature of the AI race.
The GPT-5.5 model, which 'Watermelon' is said to rival, is a formidable AI model that was released earlier this year by a competitor. The competitor has since unveiled an even more powerful model, GPT-5.6, although it has not yet been widely released due to government requests.
Previous Efforts and Future Plans
Earlier this year, the tech firm unveiled the first in a series of models known as Muse Spark. While it performed well on benchmarks, it could not match or exceed the performance of models from competitors.
The company's relentless pursuit of AI supremacy is led by the firm's founder, who has been pushing heavily for the company to lead the AI race. Last year, he appointed the current head of the superintelligence division and rebranded the company's AI division as Meta Superintelligence Labs.
Within the company, the head of the division oversees a team of top-tier AI researchers, as well as other AI initiatives, including a recent hardware push. The company has reportedly offered hundreds of millions of dollars to top AI talent in an effort to recruit them.
This aggressive recruitment comes as the company escalates its spending on infrastructure. Investors were informed this year that the firm plans to spend between $125 billion and $145 billion on chips, data centers, and other infrastructure. This marks an increase from an initial forecast of $115 billion to $135 billion, due to rising component costs and additional data center investments.