AI startup Writer launches new model to compete with OpenAI Gaylord Contreras | usagoldmines.com

San Francisco-based AI startup Author debuted a big artificial intelligence mannequin Wednesday to compete with enterprise choices from OpenAI, Anthropic and others. However, not like a few of these rivals, it would not have to spend as a lot to coach its AI.

The corporate advised CNBC it spent about $700,000 to coach its newest mannequin, together with the information and GPUs, in contrast with the hundreds of thousands of {dollars} competing startups spend to construct their very own fashions. Its technique has caught the eye of buyers.

Author is elevating as much as $200 million at a $1.9 billion valuation, based on an individual aware of the state of affairs who spoke with CNBC. That is almost quadruple the corporate’s valuation in September 2023, when it raised $100 million at a valuation of greater than $500 million.

The corporate cuts prices utilizing artificial knowledge, or knowledge created by AI. It is designed to imitate the real-world info that is normally fed into fashions with out compromising privateness and is changing into a extra in style methodology for coaching.

A examine by AI researchers revised in June discovered that if present AI growth tendencies proceed, tech corporations will “totally exhaust” the publicly out there coaching knowledge between 2026 and 2032, writing that “human-generated public textual content knowledge can’t maintain scaling past this decade.”

Amazon has used artificial knowledge in coaching Alexa, Meta has used it to fine-tune its Llama fashions, and Microsoft-backed OpenAI is incorporating it into its fashions, based on job descriptions posted by the corporate.

Some specialists, nevertheless, have warned that artificial knowledge ought to be used cautiously, because it has the potential to degrade mannequin efficiency and exacerbate current biases.

Waseem Alshikh, Author’s co-founder and CTO, advised CNBC that Author has been engaged on its artificial knowledge pipeline for years.

“There’s some confusion within the trade in regards to the definition of ‘artificial’ knowledge,” Alshikh stated. “To be clear, we do not prepare our fashions on pretend or hallucination knowledge, and we do not use a mannequin to generate random knowledge. … We take actual, factual knowledge and convert it to artificial knowledge that’s particularly structured in a clearer and cleaner method for mannequin coaching.”

The corporate’s generative AI permits company purchasers to make use of its massive language fashions, or LLMs, to generate human-sounding textual content for something from LinkedIn posts to job descriptions to mission statements, analyze and summarize knowledge or textual content and construct customized AI purposes for market evaluation and extra. The corporate has greater than 250 enterprise clients, together with Accenture, Uber, Salesforce, L’Oreal and Vanguard, who use the tech throughout sectors equivalent to assist, IT, operations, gross sales and advertising.

The generative AI market is poised to top $1 trillion in income inside a decade. Thus far in 2024, buyers have pumped $26.8 billion into 498 generative AI offers, based on PitchBook, and corporations within the sector raised $25.9 billion in 2023, up greater than 200% from 2022.