ChatGPT may be in a bit of a dilemma after its latest feature that was rolled out for Plus subscribers was recently disabled after shocking allegations came forward.
People were appalled at how the feature called Browse with Bing that the company took great pride in was actually seen navigating around paywalled articles.
Hence, the allegations have forced the leading tech giant with no other option but to disable it.
The announcement was made by OpenAI who spoke about how it takes matters like this super seriously. And therefore, it was going to implement all measures to carry out an investigation on how something of this sort can happen.
The company went on to add how it was getting reports and is yet to confirm if they’re true or not. But from what is being spoken about, it can sometimes display content in a manner that they wouldn’t like. Hence, they’re not willing to put the leading AI tool’s reputation on the line.
A recently published health update by the firm spoke about how users might sometimes put in bizarre requests like displaying the complete text from a certain URL and when that happens, their wish gets fulfilled inadvertently.
OpenAI adds how it disabled the feature starting this week and hopes to do so as a means of practicing extra care and precaution, more than anything else. They really do hope to fix this as they try to do something right for owners of the content that would be treated unfairly if things like these are let go.
In the same way, the company hopes to bring its beta version to normal at the latest date, without any form of delay. For now, OpenAI isn’t giving out any replies for requests generated by leading tech news outlets. This was especially seen after the top newsletter made reports about ChatGPT using a feature called GPT-4 browsing to gain access to online content that was guarded by paywalls.
It boldly declared how paywalled content may be attainable through prompts such as how to print the content for an article. Moreover, the next step would involve putting out URLs for the article in question after the prompt is added. This way, chatbots can highlight which contents are present in this article.
It’s a feature that could be fixed up at some point in time, the company added. And that’s obviously because it’s in violation of its own policy and related to others dealing with online publications.
Read next: ChatGPT Integration Helps Microsoft Edge Close in on 300 Million Users
People were appalled at how the feature called Browse with Bing that the company took great pride in was actually seen navigating around paywalled articles.
Hence, the allegations have forced the leading tech giant with no other option but to disable it.
The announcement was made by OpenAI who spoke about how it takes matters like this super seriously. And therefore, it was going to implement all measures to carry out an investigation on how something of this sort can happen.
The company went on to add how it was getting reports and is yet to confirm if they’re true or not. But from what is being spoken about, it can sometimes display content in a manner that they wouldn’t like. Hence, they’re not willing to put the leading AI tool’s reputation on the line.
A recently published health update by the firm spoke about how users might sometimes put in bizarre requests like displaying the complete text from a certain URL and when that happens, their wish gets fulfilled inadvertently.
OpenAI adds how it disabled the feature starting this week and hopes to do so as a means of practicing extra care and precaution, more than anything else. They really do hope to fix this as they try to do something right for owners of the content that would be treated unfairly if things like these are let go.
In the same way, the company hopes to bring its beta version to normal at the latest date, without any form of delay. For now, OpenAI isn’t giving out any replies for requests generated by leading tech news outlets. This was especially seen after the top newsletter made reports about ChatGPT using a feature called GPT-4 browsing to gain access to online content that was guarded by paywalls.
It boldly declared how paywalled content may be attainable through prompts such as how to print the content for an article. Moreover, the next step would involve putting out URLs for the article in question after the prompt is added. This way, chatbots can highlight which contents are present in this article.
It’s a feature that could be fixed up at some point in time, the company added. And that’s obviously because it’s in violation of its own policy and related to others dealing with online publications.
Read next: ChatGPT Integration Helps Microsoft Edge Close in on 300 Million Users
