Developers Look For Substitutes And Better Alternative To OpenAI

For software companies hoping to profit from an anticipated $90 billion market for artificial intelligence, Microsoft-backed OpenAI is no longer the only option availble.

More than a dozen startups and investors said they are incorporating rivals to industry leader OpenAI, casting doubt on expectations that Microsoft Corp. and OpenAI will dominate the young field. Their motivations include a desire for models tailored to specific tasks, a wariness of relying on a single company, and the chance to save money.

It is possible to predict how the next phase of generative AI, which is described as technology that can produce text, images, or other media in response to cues, would emerge from the change made by some software developers toward alternate AI foundation models.

AI investor George Mathew of Insight Partners linked the AI foundation models to past technology advances that gave rise to rivalry. Foundation models are AI systems with the capacity to learn to carry out a range of tasks that have been trained on vast datasets of data.

“Did we only have a single internet service provider?” Mathew said. “In a similar manner, we will need multiple foundational model providers for a healthy functioning ecosystem.”

He added: “The current head start that OpenAI has will not make it the only choice.”

Tome, an AI storytelling firm that makes it easier for users to create presentations quickly, was initially constructed using the GPT-3 foundation architecture, which OpenAI first released in 2020. According to Tome, it reached 3 million users this month and began experimenting with alternative models.

It has a text model from OpenAI rival Anthropic added to the mix and plans to switch from OpenAI’s photo generation model DALL-E to Stability AI’s open-source model Stable Diffusion.

According to Keith Peiris, chief executive of Tome, the objective is to identify the model that performs each operation with the least amount of time and the highest level of quality.

In an effort to deliver more dependable services, control costs, and capitalize on the specialization of several models, AI developers and investors claimed that there is a new industry consensus to lessen reliance on a single model.

Once its ChatGPT chatbot astounded many with its capacity to respond to complicated inquiries in understandable, grammatically sound, and human-appearing English, OpenAI became a household name. Microsoft has invested $10 billion in it while major competitors like Alphabet Inc.’s Google and smaller companies race to develop new models.

By many metrics, OpenAI’s recently released GPT-4 model remains the most potent.

According to PitchBook, the global market for generative AI is anticipated to reach $98.1 billion by 2026.

Foundation models, which form the foundation of AI applications, have received the most investment from strategic and venture capital investors. Players like OpenAI, which has stated that it hopes to attain $1 billion in income by 2024, depend on how these base models are employed by applications, which pay for the services.

This year, OpenAI expects to earn $200 million in income. As an illustration of how it generates revenue, it costs 6 cents to process 1,000 tokens of prompts in its most recent GPT-4 model and offers a $20/month subscription tier for ChatGPT.

Startups are also concerned that Microsoft’s use of OpenAI models across its Office Suites, search, and other products could put it in direct competition with its AI clients.

“Some of these applications will use sensitive company data, and the foundation models will see these companies’ interactions with their own customers,” said Mike Volpi, partner at Index Ventures, which backs OpenAI competitor Cohere. “Many of these companies will feel uncomfortable being dependent on Microsoft or a company generally controlled by Microsoft.”

There were no comments on the issue from OpenAI and Microsoft.

While starting with OpenAI’s models, writing assistant Jasper.ai does not want to rely on just one model, Founder Dave Rogenmoser told Reuters. It has now included Cohere and Anthropic, two other significant language model businesses with Google cloud computing relationships, and is introducing an AI engine to assist marketers in customizing voices by utilizing a variety of models.

According to CEO Matt Shumer, the AI copywriting tool HyperWrite combines each user’s activities with several models based on a range of factors. For instance, it makes use of the OpenAI model to produce lengthy articles and Cohere to more quickly and cheaply auto-complete sentences.

Simply because OpenAI is having a hard time keeping up with the growing demand, some have turned to alternatives.

“OpenAI’s servers are down a lot. We want our users to have a better experience, and using multiple models helps us to process inquiries at lower cost,” said Srinath Sridhar, CEO at Regie.ai, a writing assistant serving sales team.

Several startups, like as customer service software provider Intercom Inc., continue to support OpenAI wholeheartedly. The director of machine learning at Intercom, Fergal Reid, acknowledged that OpenAI’s GPT-4 is “quite expensive.” He continued, nevertheless, “We now believe we need to employ GPT-4 in order to obtain the accuracy level that we require for customer service.”

(Adapted from Reuters.com)

Advertisement

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s