2023-05-06 00:03:52
As every year, Appian, one of the largest software providers in the world, has held its annual conference for customers and developers in the city of San Diego. The American company has spent more than a decade being one of the biggest defenders of ‘low code’, or what is the same, the creation of applications without code or, if necessary, with very little. Thanks to this technology, it is capable of building any application in just 8 weeks.
As expected, the artificial intelligence (AI), and especially generative artificial intelligence, as is the case with ChatGPT, have had a leading role throughout the event. It is impossible to ignore the impact of these new systems, capable of generating text or images at the user’s request.
“We are facing a productivity revolution thanks to artificial intelligence and process automation,” said Matt Calkins, CEO of Appian, regarding where we are today thanks to artificial intelligence.
Appian has included several places in its app where it gives its customers access to generative artificial intelligence, but with a big distinction. One thing is public AI, what we commonly know as ChatGPT, to which everyone has access, and is trained with public data. Another is the private AIthat is, using the same open source algorithms, but trained only with the data of the company on duty to meet specific objectives for it.
Why you shouldn’t share data with ChatGPT
Los dangers of public AI for companies There are mainly four. If employees use it, they can reveal critical data to a third party, for example, to OpenAI, the company behind ChatGPT and who has access to all the conversations you have with the AI. There are also possible regulatory problems, and it is that these types of tools are in the crosshairs of regulators around the world.
Also, keep in mind that the answers offered by tools like ChatGPT may be too generic. They can also be wrong. Finally, it is normal that many companies do not want to help develop the operation of third-party tools by providing data that, in addition, may be especially sensitive for the operation of their business.
Advantages of private AI
Within the app store, Appian customers will find a connector with which they can integrate ChatGPT into their process flows, for example to summarize each incoming mail with the generative AI in a classification process. This new connector has become the fastest growing app in Appian’s history, another indicator that the ChatGPT craze is taking over across the board.
On the other hand, the company offers access to private AI through the new ‘AI Skill Designer‘, where its clients can train the same algorithms that ChatGPT uses with their own data, but internally, without this information leaving the company, and thus achieving much more specific answers, in three tasks: document classification, data extraction from a document and mail classification. The ‘AI Skill Designer’ only needs a few sample documents to learn how to extract and classify data from any other file.
Much has been said about how ChatGPT could create code, almost with the quality of a developer, but the reality is that it is in the low-code where it will probably have the most immediate effect. On the second day of the congress we were able to see some tests that the Appian team is running that seemed almost like magic, things like requesting reports with simple text, or creating a form from a PDF, were executed in an incredibly simple way.
#reasons #companies #private