The Difference Between Search Engines, Generative AI, and Automation Tools

Search engines like Google, generative AI applications like ChatGPT, and automation tools like Zapier serve distinct purposes and have different capabilities. The key is choosing the right tool for specific business needs and use cases, rather than trying to replace one with another. With an understanding of the unique capabilities of each, organizations can combine these technologies to enhance productivity.

Search Engines

Search engines, like Google, DuckDuckGo, and Bing, comb the public world wide web and add information to an index. You use an interface through a browser or app to query the index, using natural language, keywords, or Boolean searches. You receive results that take you to websites where you will look for the information that responds to your query. In some cases, you will get instant answers. Ask Google “what time is it in Oregon?” or “when will the sun set today” and you will get an answer, followed by search results.

Some search engines collect information about you to provide the best answer to your search query. Your IP address and location, previous searches, what you clicked on, and more influence the information that appears in search results.

In addition to websites, search engines crawl and index social media sites, images, videos, discussion groups and more. Results from a search appear in seconds and, depending on the query, may offer millions of results.

Search engine results are ranked by algorithms. For some, like Google, the algorithm presents results based on popularity (how many sites link to it, how many people click on it) and relevancy (how many times your keywords appear and what it knows about your language and past behaviors). Additionally, results are ranked by sponsored (paid) results and advertisements.

For most search engine results, the source of the information is apparent. It is still up to the user to assess the validity of the data and the source, as well as accuracy and currency.

Generative Artificial Intelligence

Generative AI tools are distinct from search engines in that they can generate text, images, audio/video, code, and content. Users often see an interface reminiscent of a search engine, but instead of a natural language or Boolean search query, you will enter a prompt. In a conversational way you can continue to ask questions. The AI will respond in full sentences (or whatever you request)

They can summarize and synthesize information as well. Lawyers are using generative AI tools for ediscovery, predictive analysis, drafting communications and documents, legal research, editing and review, transcription, marketing copy, and more. Far more than just ChatGPT, there are purpose built generative AI tools for almost any task. Of course, you will need to question the breadth, depth and currency of the underlying data set, how your prompts and uploads will be used, scrutinize, and verify results, and consider bias.

While machine learning applied to legal tech is not new, the more recent generative AI tools are interactive and conversational. Previous versions of conversational interfaces, like online chat bots, required a lot of human development and were limited. Platforms like OpenAI make it easier for developers to leverage the generative AI for specific purposes without needing to build the product from the ground up.

We are seeing the introduction of AI in everything from project management tools like Asana, to practice management tools like Clio to legal research tools like Vlex’s Vincent. While generative AI will influence search engine results, like Bing Chat and Bard, and enhance some automation tools, it will be a while before it replaces them.

Automation

There are a lot of automation tools, many designed for lawyers. Workflow automation tools like Zapier are used as a connector to trigger an action or sequence. Some automations are add-ons or features within existing products. The end goal of an automation tool is to reduce the number of steps it takes to accomplish something without the need for human intervention. However, a human must set up the automation initially and test it. In some cases getting the automation set up requires time and skills.

A specific use case for automation is document automation, with tools like HotDocs, Lawyaw, Gavel, the Form Tool, Woodpecker, and more turning document templates into fillable “smart” documents that can apply logic, clauses, calculations, and more. Some will even let you generate client interviews that are converted to final documents. Tools built into Microsoft Word natively will let you merge data from a form into a document, as will many practice management applications.

Some examples for lawyers are rules-based calendar tools that create ticklers and reminders based on court rules on a firm calendar. Some practice management applications have workflows that are engaged based on user activity. Many client relationship management tools can automate client intake based on a few inputs. The end user must identify and enable the workflows and test that they continue to work properly.

Automation tools like Zapier, IFTTT, Microsoft Power Automate and more let users create triggers that ignite an action, often between two different applications. For instance, you can use Microsoft Power Automate to create an Outlook appointment based on an Asana deadline. Or move an email attachment from Outlook to Microsoft Teams based on certain criteria. Or trigger a request for approval if a specific person edits a document. In Zapier you will find even more options, with no code automation across over 6,000 apps. You can get ideas by industry, by app, by company size and by role. While tools like Zapier are incorporating AI tools like chatbots and conversational interfaces, they are not generative AI. It is important to test your “zaps” and automations, pay attention to error messages, and do some reality checks to make sure your automations are working.  As always, when you begin using a third-party tool and allow for API access to your existing products, make sure you understand the implications and access.

Conclusion

In summary, these tools complement each other and can be used in combination to improve efficiency and productivity in law firms and small businesses. Google helps with research and general information retrieval, generative AI enhances text-based tasks, and automation tools like Zapier optimize workflows and save time. The choice of tool depends on the specific needs and objectives of the organization.

This blog was written by Catherine Reach and originally published on From the Center Blog. Access more from the Center for Practice Management.