Your Tech Story

ChatGPT

ChatGPT Creator OpenAI Is Testing Content Moderation Systems

ChatGPT Creator OpenAI Is Testing Content Moderation Systems

Leave a Comment / Artificial Intellegence, News, Tech News / By

One of the trickiest problems on the web for years has been content management. Given how subjective it is involved in deciding what information should be allowed on a certain platform, it is a challenging topic for anybody to approach. OpenAI, the company that created ChatGPT, believes it can be of assistance and has been testing GPT-4’s capacity for content moderation. To develop a content management platform that is scalable, uniform, as well as customizable, it utilises the big multimodal model.

ChatGPT Creator OpenAI Is Testing Content Moderation Systems
Image Source: thestar.com

In a blog post, the business claimed that GPT-4 can, not solely assist in content moderation choices, but also in the development of rules and the speedy iteration of modifications to existing policies, lowering the process’s duration from months to hours.

It proclaims that the model can quickly adjust to any adjustments and interpret the numerous rules and nuanced aspects of content restrictions. According to OpenAI, this leads to more consistently labelled material.

OpenAI asserts GPT-4 moderating technologies enable businesses to complete about six months’ worth of work in a single day.

“We believe this offers a more positive vision of the future of digital platforms, where AI can help moderate online traffic according to platform-specific policy and relieve the mental burden of a large number of human moderators, ” OpenAI’s Lilian Weng, Vik Goel and Andrea Vallone wrote. “Anyone with OpenAI API access can implement this approach to create their own AI-assisted moderation system.”

Source: engadget.com

Manually assessing distressing information may have a major negative effect on human reviewers’ mental health, especially when it involves graphic content, as has been thoroughly documented. More than eleven thousand moderators will get a minimum of one thousand dollars in compensation from Meta in 2020 for any mental health problems that may have resulted from analysing content that was uploaded on Facebook.

Also Read:  Amazon Is Imposing Fee on Sellers Who Ship Products Themselves

“Judgments by language models are vulnerable to undesired biases that might have been introduced into the model during training. As with any AI application, results and output will need to be carefully monitored, validated and refined by maintaining humans in the loop,” OpenAI’s blog post reads.

Source: engadget.com

AI algorithms are not flawless. Major businesses have long used artificial intelligence in their moderation procedures, yet even with the help of technological advances, they frequently make poor content selections. It will be interesting to see if OpenAI’s technology will prevent many of the big moderation pitfalls that we have observed other businesses making throughout the years.

Chatgpt

Why is Congress limiting the use of Chatgpt in the offices?

Leave a Comment / Story, Tech Story / By

According to a recent report from Axios, the House of Representatives has introduced new regulations concerning the utilization of ChatGPT in congressional offices.

The report states that non-ChatGPT chatbots are currently not authorized for use. In a memorandum released on Monday, House Chief Administrative Chief Catherine L. Szpindor specified that lawmakers and staff are now limited to employing ChatGPT Plus, the paid version of the chatbot, due to its advanced privacy features.

chatgpt
Image Source: tech.co

The usage of the product is strictly confined to “research and evaluation” purposes, with privacy settings enabled. Additionally, offices are prohibited from inputting any non-public text into the service.

Several private companies, including industry giants like Samsung and Apple, have already imposed restrictions or outright bans on the usage of generative AI tools such as ChatGPT. These companies have expressed concerns about the potential leakage of confidential data through these tools.

These concerns are reinforced by previous privacy mishaps by OpenAI, the creator of ChatGPT, such as a bug that temporarily exposed chat histories to other users.

Also Read: Can the iPhone 16 be cheaper than the iPhone 15 Pro?

In the memorandum issued on Monday, Szpindor explicitly stated that no other versions of ChatGPT or similar AI software based on large language models are currently authorized for use in the House.

This announcement follows Senate Majority Leader Chuck Schumer’s recent call for Congress to expedite the enactment of new legislation aimed at regulating the artificial intelligence industry. Schumer has presented a framework that outlines Congress’ focus areas, including the potential risks of AI to national security and employment.

Schumer emphasized the significance of AI as an innovative force that can usher in technological advancements, scientific discoveries, and industrial growth. He stressed the need to encourage innovation while ensuring its safety. Failure to provide a sense of security around AI development could hinder progress in this field.

Lawmakers in both the Senate and the House have already introduced several bills aimed at regulating the AI industry this year. Sens. Richard Blumenthal (D-CT) and Josh Hawley (R-MO) recently proposed a measure that seeks to exclude AI companies from the protections of Section 230, thereby exposing these companies and their products to legal liability.

Furthermore, Congress has initiated discussions with prominent figures from the industry. Sam Altman, CEO of OpenAI and creator of ChatGPT, received a warm welcome from the Senate in May.

Other proposed measures include requesting federal agencies to review their AI policies and establishing a new commission tasked with studying and establishing regulations for the industry.

ChatGPT

OpenAI’s ChatGPT app tops 500K downloads in just 6 days

Leave a Comment / News, Tech News / By

The ChatGPT app from OpenAI has had a fantastic start despite only being available on iOS and in the United States until today’s extension to 11 more international regions.

In its first six days of availability, the app has already topped half a million downloads, according to a recent report by app intelligence firm data.ai. It is only surpassed by the debut of the Trump-backed Twitter clone, Truth Social, in February 2022 in terms of new app releases that have had the highest performance over the course of this year and the previous year.

ChatGPT
Image Source: techcrunch.com

The App Store has become flooded with third-party apps identifying themselves as “ChatGPT” or “AI chatbot” as customer demand for AI chatbots increased.

Despite the fact that many of these were really “fleece are,” aiming to deceive users into paying for pricey memberships to access their AI, a collection of top applications nevertheless managed to generate millions in revenue from users.

It might have been more difficult for an official ChatGPT app to gain popularity given the competitive environment among AI chatbots. In actuality, though, such was not the situation.

Also Read: Windows 11 finally gets native RAR support

The majority of its competitors, including other well-known AI and chatbot apps as well as Microsoft’s Bing and Edge apps, which provided some of the first important third-party integrations of OpenAI’s GPT-4 technology, were outperformed by the ChatGPT app from OpenAI.

Though Bing and Microsoft Edge undoubtedly benefited from ChatGPT’s initial popularity, they saw 340,000 and 335,000 downloads on iOS and Android, respectively, in their best five-day periods in February. However, OpenAI’s ChatGPT app easily outperformed them, producing 480,000 installs in the initial five days of its United States launch, when the app was iOS-only

The app outperformed other top AI chatbot apps in the U.S., many of which had generic names to take advantage of consumer searches for terms like “AI” and “chatbot” on the App Store, according to Data.ai’s data. When compared to other apps’ best five-day periods throughout the App Store and Google Play in 2023, OpenAI’s ChatGPT came in at number five in terms of downloads.

According to the data, “Chat with Ask AI,” the sole app that outperformed it, got 590,000 installs between April 4 and 8, 2023, as opposed to ChatGPT’s 480,000 downloads between May 18 and 22. Even though it has only been launched for a week, ChatGPT is already among the top five AI chatbot apps in terms of downloads in May 2023 in the United States.

OpenAI

OpenAI to introduce ChatGPT app for iOS

Leave a Comment / News, Tech News / By

The ChatGPT app is going mobile. After the App Store was flooded with questionable, unofficial services for several months, OpenAI announced the release of an official iOS app that enables users to use its well-known AI chatbot while on the road.

The company claims that the newly released ChatGPT app will be free to download and use, ad-free, and voice-enabled, but it will only be available to users in the United States at first.

OpenAI
Image Source: freemalaysiatoday.com

The ChatGPT app, like its desktop equivalent, enables users to communicate with an AI chatbot to ask queries without performing a standard online search and to receive advice, discover inspiration, study, conduct research, and do other things.

Also Read: Google to delete inactive accounts starting December

Given the problems with Apple’s own voice assistant, Siri, and Apple’s own lack of advancement in artificial intelligence, the latest update may encourage more customers to adopt ChatGPT as their primary mobile assistant.

Since Google now reaps the benefits of serving as the default search provider in Safari on Apple’s iPhone, the launch may also have an effect on Google.

When using ChatGPT on a mobile device, the app will sync your history across devices, which means it will remember any searches you’ve made through its web interface and make those results available to you.

In order to support voice input, Whisper, an open-source speech recognition technology from OpenAI, is also integrated into the app.

OpenAI claims that ChatGPT Plus members will have early access to additional features and quicker response times in addition to being able to use GPT-4’s capabilities through the new app. “ChatGPT Plus subscribers get exclusive access to GPT-4’s capabilities, early access to features, and faster response times, all on iOS,” the company stated.

The technology that underpins the highly successful ChatGPT has been replaced by the potent artificial intelligence model GPT-4. Earlier in February, OpenAI unveiled a $20 per month test subscription package for their well-known AI-powered chatbot, named ChatGPT Plus.

Also Read: WhatsApp Lock Feature to Protect ‘Extra Special’ Chats?

According to the corporation, the roll-out of iOS is already underway in the United States and will soon be extended to new nations. With the release of ChatGPT at the end of last year, the public competition to control the AI technology market began. This spurred big giants like Alphabet Inc. and Meta to promote their own products.

The release of the ChatGPT app coincides with major tech companies, such as Google, Microsoft, and Facebook, experimenting with AI. Google and Microsoft have also integrated artificial intelligence (AI) features into their search engines, the latter of which was accomplished through a costly partnership with OpenAI.

But having direct access to ChatGPT on a mobile device rather than through a search engine or browser may significantly alter how users now use their phones to look for and engage with information.

ChatGPT

Italy Restores ChatGPT after OpenAI Responds to Regulator

Leave a Comment / News, Tech News / By

According to the agency and the corporation, the ChatGPT chatbot has been reinstated in Italy after OpenAI resolved concerns expressed by the country’s data protection body.

After the Italian data protection authority, designated as Garante, temporarily suspended the chatbot and opened an investigation into the artificial intelligence program’s alleged violation of privacy laws, OpenAI powered by Microsoft, banned ChatGPT in Italy this past month.

ChatGPT
Image Source: satlokexpress.com

Garante had granted OpenAI till Sunday for it to alleviate its worries before permitting the chatbot to resume operations in the nation.

Garante claimed a month ago that ChatGPT didn’t have any legal justification for the huge collection and storage of users’ personal data required to train the chatbot.

Also Read: OpenAI rolls out ‘incognito mode’ on ChatGPT

Garante had also charged OpenAI with failing to verify the legal age of ChatGPT users, who are required to be 13 or older. In response, OpenAI announced it will provide a tool to confirm the age of users in Italy at the time of registration.

The firm announced on Friday that it would make its privacy policy as well as input from users’ opt-out form more visible.

According to a company spokesperson, it will also make available an enhanced way for users in the European Union to take advantage of their privilege to protest its utilization of private information to train its models.

Individuals who wish to opt-out must fill out an elaborate form with their personal information, including any proof of data processing via pertinent prompts.

Garante expressed its appreciation for the measures made to balance technical advancement with adherence to human rights and expressed the hope that the firm would continue on this road toward achieving compliance with European data security standards.

Although ChatGPT’s swift growth has drawn the interest of lawmakers as well as regulators in multiple nations, Italy was the first Western European nation to restrict it.

Also Read: How Will ChatGPT Change Education and Teaching?

On Thursday, a panel of EU parliamentarians approved new regulations requiring companies using generative AI tools, including ChatGPT, to declare any copyrighted data used to create their systems.

The organization that unifies Europe’s national privacy regulators, the European Data Protection Board, established an investigation force on the chatbot previously this month in response to Garante’s concern regarding ChatGPT.

Garante stated that it will cooperate with the special task team and carry out its investigation into ChatGPT.

OpenAI

OpenAI rolls out ‘incognito mode’ on ChatGPT

Leave a Comment / News, Tech News / By

OpenAI declared on April 25 that it is creating what one employee referred to as an “incognito mode” for its popular chatbot ChatGPT which does not record people’s chat histories or use them to advance its artificial intelligence.

Additionally, the business which is based in San Francisco announced plans for a “ChatGPT Business” subscription featuring more extensive data management.

OpenAI
Image Source: firstpost.com

The decision was made as concern has grown about how ChatGPT along with the other chatbots it influenced handle the data from numerous billions of users, which is frequently utilized for “training” or improving artificial intelligence.

Italy this month suspended ChatGPT due to potential privacy concerns, but said OpenAI might reinstate the service provided certain conditions were met, which include providing users with the ability to reject the collection of their data. France, as well as Spain, also initiated inquiries into the service.

Mira Murati, Chief Technology Officer (CTO), OpenAI, explained to Reuters that the business complies with European privacy laws and is striving to reassure regulators.

Also Read: Google Authenticator finally syncs one-time codes in the cloud

She said that the updated functions were the result of months of effort aimed at putting people “in the driver’s seat” concerning data collecting rather than Italy’s ChatGPT ban.

“We’ll be moving more and more in this direction of prioritizing user privacy,” Murati said, with the goal that “it’s completely eyes off and the models are super aligned: they do the things that you want to do”.

Source: tech.hindustantimes.com

She stated that user data has assisted OpenAI in improving the software’s dependability and reducing political bias, amongst other problems, but added that the business still has difficulties.

With the product update from Tuesday, users can export their data and turn off the “Chat History & Training” option within the settings.

Users can now deactivate “Chat History & Training” in their options and export their data because of its additional features. However, before totally wiping the discussions, OpenAI will keep them on file for 30 days to monitor for any potential abuse. With the upcoming “ChatGPT Business” subscription, chats won’t by default be utilized for training artificial intelligence models.

The business will nevertheless keep conversations for about 30 days to watch for any kind of abuse before eliminating them, according to Nicholas Turley, the OpenAI product officer, who compared this to the incognito mode of an internet browser.

Additionally, the business subscription offered by the company won’t by default use chats for training artificial intelligence models when it becomes accessible in the coming months.

Businesses already have access to ChatGPT thanks to Microsoft Corp.’s investment in OpenAI. The existing clients of the cloud provider, according to Murati, would be interested in that service.