A Brief History of AI
While most people are more interested in the progress made in AI technology over the past six months, understanding its history will provide a little bit of context. In the 1950s, AI was referred to as a deterministic technology (basically a big series of if-thens). Fast-forward to today and AI is now considered more as a probabilistic model that uses large sets of data to statistically predict the next node in a series based on probability. More recently, things with AI have begun to speed up. In 2016, Google declared themselves an "AI-first company," literally inventing the 'T' (or Transformer technology) in “GPT”, that led to big advances. Fast-forward to the beginning of 2023, when OpenAI takes its GPT technology and wraps it in a chatbot and creates chatGPT. That gets us to the present – well, almost.
In the subsequent months, Microsoft began integrating OpenAI's GPT technology into its search and office suite of products, with other large tech companies following suit. Google released its own generative chat technology called Bard and subsequently rolled out additional AI offerings including PaLM 2. Amazon also released Bedrock on top of its existing suite of AI tools that includes SageMaker and others. For many companies, this means that their existing technical architecture may have an extension of AI services that is currently (or will soon be) available for implementation.