OpenAI, Mistral, And Google Launches Updated AI Models

by

The rivalry among companies in the Artificial Intelligence (AI) sphere, including OpenAI, Mistral, and Google, is heating up as they explore innovative ways to apply AI technology.

Updated AI Models From OpenAI Entering The Market

Within a day’s time, OpenAI led by Sam Altman, Google, and the French AI company Mistral each unveiled new upgrades to their AI systems.

As the industry gears up for the unveiling of the latest edition of OpenAI’s GPT engine, the excitement builds following hints dropped by Facebook’s Global Affairs President Nick Clegg about an upcoming enhancement to Meta’s AI model, Llama.

At an London event, Clegg notably discussed the introduction of Llama’s third iteration, known as Llama 3.0. Shortly following this announcement, Google unveiled Gemini Pro 1.5 – a model they had been working on for some time. It’s important to mention that prior to the Pro version, Google introduced an upgraded Gemini 1.5, which was marketed as a more efficient alternative to Google 1.0 Ultra, capable of helping users accomplish numerous tasks while requiring fewer computing resources.

Currently, Gemini Pro 1.5 is accessible to the general public, but its free plan can handle a maximum of 50 requests per day. Shortly after its release, OpenAI introduced an advanced model – the last version of GPT-4 Turbo, named Turbo.

A significant achievement: The leading AI firm announced that the GPT-4 LLM version unveiled in mid-March represents their most sophisticated Large Language Model to date.

Gemini Pro 1.5 and GPT-4 Turbo are versatile systems that support various modalities, allowing them to process text, spoken words, and videos.

Concerns Raised About Mistral’s Launch Approach

Lastly, Mistral unveiled its innovative model, named Mixtral 8x22B. In contrast to its American peers who openly share their tools through an open-source platform, Mistral opted for a more traditional method to introduce Mixtral 8x22B. Users could access the massive 281GB file via a straightforward download link. This approach has sparked controversy among various entities due to concerns over potential risks.

Using simple terms, this method is thought to give developers no control to prevent their systems from being utilized for negative purposes once implemented.

At the same time, Intel and Goldman Sachs are now exploring advanced models to challenge AI offerings from NVIDIA and OpenAI.

Read More

2024-04-10 21:46