A few months back, we saw tech giant OpenAI justify how its ChatGPT is capable of carrying out political disinformation initiatives and therefore, it was making changes in its respective policy to prevent that from taking place.
Everything seemed to be under control or it appeared so but now, a new article from The Washington Post is shedding light on how that might not be the case. And perhaps we should be more concerned with how the company is behaving in this regard.
Think along the lines of the chatbot being used to break all the designated rules in place, producing the worst repercussions on the upcoming US elections.
The company’s user policies are designed to carry out a ban on all kinds of political campaigns. So if you’d like to produce a huge quantity of materials for your next political campaign, you might wish to think again. As such materials generated in bigger volumes than usual wouldn’t be allowed, including projects designed to target a particular kind of demographics, not to mention create chatbots to spread information that might not have the greatest amount of truth affiliated with it. Other than that, lobbying and even politically-themed advocacy wouldn’t get a thumbs up either.
Speaking with Semaor, the parent firm of ChatGPT says that it’s got zero room for this kind of action. Therefore, it was now working on designing the best classifier for machine learning that would report to the firm, whenever huge texts pop up regarding election campaigns or such related actions.
But no matter how much reassurance is generated in this regard, the company seems to be failing in terms of actually bringing all of its efforts to light. Moreover, in the past couple of months, we’ve seen an investigation come to light by the Washington Post regarding how one can simply add in a prompt such as motivating women from the suburban regions to come forward and vote for a certain presidential candidate.
This includes enlisting down any kind of policies that may benefit the younger voting generation in particular. So when you actually put things into perspective, in the past, OpenAI was all about tackling the great political risks but now, it’s more linked to steering clear of that kind of area.
The firm’s spokesperson did ensure that their goal is also to prevent blockage of campaigns or material that it feels could be helpful to some audiences and actually aren’t violent in nature. For instance, campaigns that are designed to stop diseases from spreading or perhaps those linked to the marketing of helpful businesses that are functioning on a small scale.
Similar to how social media apps came about, the company and its initial startup are really having trouble in terms of tackling the many issues linked to moderation. The only difference right now is how it’s not linked to just shared content but also who needs access to such tools and what the conditions for it too.
OpenAI did reveal this past month how it’s putting out a new kind of system for content moderation that can not only be customized but sized and function consistently as well.
In the past year, the efforts in terms of the app’s regulation aren’t exactly where experts had hoped they would be. But with time, they seem to be picking up the pace as lawmakers introduce new laws that would prevent the production of such works from receiving a certain kind of protection under a certain law.
Read next: Musk, Zuckerberg, and More to Converge with Chuck Schumer's AI Extravaganza
Everything seemed to be under control or it appeared so but now, a new article from The Washington Post is shedding light on how that might not be the case. And perhaps we should be more concerned with how the company is behaving in this regard.
Think along the lines of the chatbot being used to break all the designated rules in place, producing the worst repercussions on the upcoming US elections.
The company’s user policies are designed to carry out a ban on all kinds of political campaigns. So if you’d like to produce a huge quantity of materials for your next political campaign, you might wish to think again. As such materials generated in bigger volumes than usual wouldn’t be allowed, including projects designed to target a particular kind of demographics, not to mention create chatbots to spread information that might not have the greatest amount of truth affiliated with it. Other than that, lobbying and even politically-themed advocacy wouldn’t get a thumbs up either.
Speaking with Semaor, the parent firm of ChatGPT says that it’s got zero room for this kind of action. Therefore, it was now working on designing the best classifier for machine learning that would report to the firm, whenever huge texts pop up regarding election campaigns or such related actions.
But no matter how much reassurance is generated in this regard, the company seems to be failing in terms of actually bringing all of its efforts to light. Moreover, in the past couple of months, we’ve seen an investigation come to light by the Washington Post regarding how one can simply add in a prompt such as motivating women from the suburban regions to come forward and vote for a certain presidential candidate.
This includes enlisting down any kind of policies that may benefit the younger voting generation in particular. So when you actually put things into perspective, in the past, OpenAI was all about tackling the great political risks but now, it’s more linked to steering clear of that kind of area.
The firm’s spokesperson did ensure that their goal is also to prevent blockage of campaigns or material that it feels could be helpful to some audiences and actually aren’t violent in nature. For instance, campaigns that are designed to stop diseases from spreading or perhaps those linked to the marketing of helpful businesses that are functioning on a small scale.
Similar to how social media apps came about, the company and its initial startup are really having trouble in terms of tackling the many issues linked to moderation. The only difference right now is how it’s not linked to just shared content but also who needs access to such tools and what the conditions for it too.
OpenAI did reveal this past month how it’s putting out a new kind of system for content moderation that can not only be customized but sized and function consistently as well.
In the past year, the efforts in terms of the app’s regulation aren’t exactly where experts had hoped they would be. But with time, they seem to be picking up the pace as lawmakers introduce new laws that would prevent the production of such works from receiving a certain kind of protection under a certain law.
Read next: Musk, Zuckerberg, and More to Converge with Chuck Schumer's AI Extravaganza
