You can test this yourself by opening a patent application you've written in Microsoft Word and then using the keyword search tool. The keyword search will easily find any keyword you want, but it cannot find ideas. For example, try asking the keyword search to find if the document discusses “safety mechanisms.” Unless the document includes the words “safety mechanisms,” the keyword search will not return any results even if the document includes many different types of safety mechanisms.
This simplistic analogy explains why keyword-based prior art searches do not work well because they search for keywords rather than searching for ideas (which can be expressed by countless different words). In contrast, our advanced AI understands the idea you're searching for and understands which patents are relevant to that idea (even when the patents use a wide range of diverse and even surprising words).
Imagine your client excitedly tells you about a bike brake invention. You then conduct a search for “bike and brake” only to discover there are over 18,000 U.S. patents and publications. To narrow the results, you search for “bike and brake and cantilever,” but then you realize there are countless different words that previous patent drafters could have used to describe similar bike brakes. As a result, (1) any keywords you choose will omit applicable patents, and (2) you need to read thousands of patents to really know which patents are relevant.
Let's say your client wants to patent the following idea:
A drone that detects and then fights wildfires. The drone uses infrared cameras to find and analyze fires. The drone sprays fire suppressant to extinguish small fires.
A keyword search might look like the following: “drone wildfire camera extinguish.” The problem with this old-fashioned keyword search is that it's a jumble of unrelated words that will (1) return many irrelevant patents and (2) fail to return the most applicable patents.
In contrast, our advanced AI understands complete ideas and then carefully analyzes the prior art to find those ideas even when they are written in different ways. The result is a far more thorough search followed by a much more focused list of results.
For all the reasons described above, even if you prefer to outsource your prior art searches to human searchers, please make sure they are using our advanced AI rather than just keyword searches.
