How to get ChatGPT to stop writing so much nonsense

code in mind

ChatGPT suffers from hallucinations, which means: The AI ​​puts together texts that are fantasized together from different sources.

(Photo: Moment/Getty Images)

Dusseldorf ChatGPT is like an intern with over-confidence: the program completes tasks on the fly and presents the results in a tone as if it had written a thesis on the subject. However, when professionals scrutinize the work closely, they often discover inaccuracies and errors. This limits the possible uses.

Despite these weaknesses, the latest generation of artificial intelligence (AI) on which ChatGPT is based offers great potential for business. Large language models can formulate e-mails, summarize business reports, answer customer complaints – and much more.

But as with an intern, the following applies: if it is really important, an experienced person must always check the result afterwards.

The productive power of these programs can only really be used if AIs like ChatGPT are able to get rid of their fantasies. It’s about turning the exaggeratedly self-confident intern into someone with a realistic self-assessment, who will let you know when he doesn’t know what to do anymore.

Read on now

Get access to this and every other article in the

Web and in our app free of charge for 4 weeks.

Further

Read on now

Get access to this and every other article in the

web and in our app.

Further

source site-12