6 Business Applications that Badly Need Better AI
The success and growth of AI is undeniable. Yet there are still basic tasks performing poorly, despite or because of automation. In some cases, you can blame reliance on outdated AI. In other cases, it is a result of corporate policies or multiple AI systems that compete against each other. The AI systems in question may be top notch, but they play against each other. It is similar to a Wall Street firm using two separate, competing black box trading systems, resulting in cannibalism. Each system may be doing great, but combined together, they annihilate each other.
It is sometimes said that dentists have the worst teeth. Perhaps, some of the companies with the best AI teams have the worst AI systems?
1. Google Search
Google brags about its new neural network with 500 billion parameters (see here), yet its search engine got worse over time, not better. Search results, on average, returns only basic answers to your questions these days. A fact that everyone is familiar with. It is a great tool for people with below average intelligence. If you are looking for mathematical or technical articles, especially recent ones, you need to use alternate search tools, like StackExchange or arXiv.com. Or include 2022 and arXiv as extra keywords in your search query.
Many new, high quality content is not even indexed by Google […]
Read full article here.