A New Way for Tech Giants to Make Money: Small Language Models

wallstreetcn ·  May 19 16:15

Source: Wall Street News

After spending billions of dollars to develop big models, tech giants are turning their attention to small models, hoping to use this to open up a new path of profit.

According to media reports on Saturday, big models were once highly anticipated and are seen as a key driving force for the development of generative AI. However, restrictions such as high operating costs and data privacy concerns have hindered the widespread application of this technology in the enterprise field.

Today,$Apple (AAPL.US)$,$Microsoft (MSFT.US)$,$Meta Platforms (META.US)$und$Alphabet-A (GOOGL.US)$/$Alphabet-C (GOOG.US)$Other tech giants have launched small models with few parameters but strong performance, with the intention of opening up AI profit growth points.

Generally speaking, the larger the parameters, the more powerful the model. However, large parameter models often require massive data and computing power support. Their operating costs and energy consumption are high, which is undoubtedly a heavy burden for small and medium-sized enterprises.

However, small models are popular for their advantages such as relative cheapness, energy efficiency, and customization.

Eric Boyd, corporate vice president of Microsoft's Azure AI platform, said that the small model opens up more applications for customers that were previously impossible due to insufficient return on investment.

The “Gemini Nano” small model released by Google this week has been built into the new Pixel phone. Meta's new 800 million parameter Llama 3 claims to be comparable to OpenAI's GPT-4 model, and French startup Mistral has joined this camp.

At the same time, the highlight of the small models is that they can ensure data privacy. These models can complete operations locally on the device, and there is no need to upload sensitive information to the cloud, making them attractive to enterprise customers who pay attention to data confidentiality.

Furthermore, the smaller models have relatively low energy consumption, which can drive the extension of artificial intelligence from the cloud to terminal devices.

Apple released a small OpenELM model last month, hinting at the intention of incorporating AI features into the best-selling iPhone series.


The translation is provided by third-party software.

The above content is for informational or educational purposes only and does not constitute any investment advice related to Futu. Although we strive to ensure the truthfulness, accuracy, and originality of all such content, we cannot guarantee it.
    Write a comment