Google's AI Efforts After ChatGPT

Google’s AI Efforts After ChatGPT


The tech world has recently focused on artificial intelligence – primarily with ChatGPT.  Microsoft has funded the creator, OpenAI who made it available to the public.  The chatbot AI system has become a sensation in the few months since its release in November 2022.

ChatGPT works through conversations and can provide answers, determine its mistakes, ask clarifying questions, and identify relevant requests. Several applications have already been developed based on its API to create content.

In response to the attention ChatGPT has garnered, Google’s parent company Alphabet, Inc., announced the development and availability of their own conversational AI solution – Bard.

Bard, like ChatGPT, is powered by a neural network to accomplish its conversational interface.  Google refers to this as Language Model for Dialogue Applications (LaMDA).  This is a series of algorithms that are used to identify relationships in a dataset. In layman’s terms, it’s a model that simulates how human brains work.

At this time, Bard will only be available to a limited audience for testing.  Google has long made AI a part of its search system to help find relevant information for user queries.

Google Products Already Using AI Solutions

Improvements to its search system have already been implemented in a few different ways.  These include search improvements using images and self-updating maps in Google Maps.  There is also a lot of progress in preserving languages and culture.

The implementation of Google Lens in the Google app on mobile devices uses AI to assist in finding a relevant result based on the image selected.  

Google Lens on an iPhone

Here Google’s search engine can search through specific libraries.  For example, if you’re looking at a plant, it will analyze the image to look for plant characteristics to help find its identification.

It also works to identify text in an image and provide translations.  Google Lens isolates the text in an image and then provides a translation if needed. In the example below, you can see the translation overlaid on the right image.

Twenty peso bill

Google Partnerships

Google is working with several partners in their efforts to develop AI solutions. These partners include Cohere, C3, and Anthropic.

Cohere is using Google Cloud to speed up language model training implemented in chatbots, content moderation, and categorization. Their efforts have resulted in advances in training efficiency.

C3 uses AI through Google to improve supply chain capabilities, manufacturing, optimization, and predictive maintenance. Their use of Google products includes Google Cloud, Google Manufacturing Data Engine, Vertex AI, and Cortex framework.

Anthropic primarily focuses on research that can lead to reliable, interpretable, and steerable AI systems.  Their efforts include research to use AI for natural language, reinforcement learning, and code generation.

OpenAI’s ChatGPT has a headstart with its public release, but Google has already been in progress with its own efforts at adding AI to its products and partnership with other contributors.  This competition is a good sign that the industry is in good health, with many projects in development. 

The future of AI may have computers creating content on their own, but the assistance of humans is still needed to add value to the end product. Even with today’s advances, AI still has a long way to go before fully integrating into our daily lives.



Source link