This is a topic that is trending at the moment and may change how we use the web in future. With the development and expansion of large language models into all kinds of creative spaces. For the previous 20 or so years people have been using the web to search for relevant content on web pages from around the world. Now it would seem that we can ask a large language model and it can produce vast amounts of information in far more detail than can be found on a web page. Mainly due to the advances in Natural Language Processing that have taken place of late. As illustrated by OpenAi's ChatGpt, Google's Gemini, Anthropic Claude, Microsoft's Co-pilot to give some examples. However there are many many more being developed all the time. With more new ai tools with an ever expanding use cases being experimented on created every week.
It is also changing how to code, making the ability to create simple programs easier and quicker. Experiments
The revolution of the web from search, read and consume to a more interactive where we chat with chatbots and they help us in ways we never thought possible is very much its beginning phase.
Machine Learning is very much an iterative process where it learns from its mistakes and improves to be ever more helpful is very much on its way. This is very much the more intriguing aspect of the new wave of artificial intelligence models. Its ability to learn and adapt. Rather than previous coding models where the code was fixed and coded by humans. Where changes would be gradual. These new coding models allow far complex changes to the code to happen a lot quicker. One of the problems with previous coding models was that the code could get so large, that to change it and ensure it worked without errors or bugs would take a massive team immense amounts of time. If the computer itself could ensure the code was correct and without errors or bugs, then any changes made could be done a lot easier and quicker. The human is still very much in control but the adaptability of the code to meet the needs of individual coders is far more useful. Requiring far less human resources allowing coders to be less focused on the code itself but on the results of the code produced.
How this will change things is up for debate. How it will change society we don't know. There is always those resistant to change while others try to navigate through the change, Others try to control the change but I don't think such an approach would work as the training and learning algorithms iterate over time. It will just get to a Plato of usefulness then grow over from there,
The machine learning nature of the tools means they will inevitably learn and become more useful. Many professionals such as professionals that use them in medical procedures use machine learning to help fight cancers and other complicated diseases. Able to highlight known patterns as well as un-usual patterns. They use a percentage to evaluate the accuracy of their responses.
One thing that is obvious from the major companies that are working on these tools. Is that each are concentrating on different set of criteria. Claude is very much focused on programming which is a unique take on the LLM's.
Grok, is developing fast and is contemporary in feel.
ChatGpt is trying to be thorough.
Gemini is trying to be everything to everyone.