The Telegram CEO who was recently arrested in France back in August was charged on the grounds of enabling criminal activity on Telegram. Some of these illegal activities include money laundering, cyberbullying, and the circulation of deep fake porn and Child Sexual Abuse Material (CSAM). Drug trafficking and fraud also took place on the platform but Telegram turned a blind eye to these activities.

Despite government pressure to moderate content and disclose data, Durov remained his ground in the name of free speech. The French President, Emmanuel Macron even had to put out a statement regarding the nature of his arrest. Macron claimed that Durov’s arrest was not for political reasons.

New Telegram Features: What the new features do

Taking into account everything that has happened in the past few weeks, Durov has stated that Telegram will be bringing in new features. These new Telegram features come as a result of the criticisms and backlash the platform has received over the lack of content moderation.

Features that had been previously known to be abused by anonymous and malicious users have been disabled. New media uploads to the Telegraph standalone blogging tool have thus been disabled. This feature was once notorious for being misused by anonymous actors.

Telegram has also removed features that are not as commonly used by its user base. For instance, the app has removed the People Nearby feature which has been known to contain issues with scammers as well as bots. Business Nearby has instead taken its place. This new feature shows verified and legit businesses.

Telegram will be turning a new leaf and Durov’s arrest is the glaring catalyst for this new chapter for Telegram.

Suggested:

Telegram Stars Enables Content Monetisation and Digital Purchases.

Username and PIN: WhatsApp’s Upcoming Privacy Feature.

The post New Telegram Features Rolling Out After CEO’s Arrest appeared first on Tricky Enough.

To lessen the mental load on human moderators, OpenAI unveiled an AI moderation system based on GPT-4. In order to help with content filtering on internet platforms, OpenAI recently disclosed that it had created an AI system employing GPT-4.

In comparison to conventional human-led moderation. The company claims that this technology enables faster iteration of policy changes and more uniform content labeling.

With this change, we hope to increase policy updates’ speed, improve consistency in content labeling, and lessen our reliance on human moderators. It could also improve the mental health of human moderators, demonstrating the potential for AI to protect mental health online.

Moderation of Content Issues

Content moderation, according to OpenAI, is difficult work that demands painstaking attention. A nuanced comprehension of context, and constant adaptability to new use cases.

Traditionally, human moderators have been responsible for these time-consuming responsibilities. They examine a lot of user-generated content and eliminate any hazardous or unsuitable information.

This type of labor may be psychologically exhausting. The human cost of online content moderation may be decreased by using AI to perform the task.

How Does the AI System at OpenAI Work?

By utilizing GPT-4 to read content policies and reach moderation decisions, OpenAI’s new approach intends to aid human moderators.

Experts in the policy first draught label samples and content standards that follow the policy. After that, GPT-4 identifies identical examples without revealing the reviewer’s responses.

GPT-4’s labels can be improved by comparing them to human labels, and OpenAI can retrain the AI until it consistently understands the rules.

Suggested:

The Most Recent ChatGPT iOS App Update Includes Customised Instructions.

US Workplaces Experience ChatGPT Craze, Raising the Alarm For Some.

The post Is It Possible For GPT-4 To Lower The Human Cost Of Content Moderation? According to OpenAI appeared first on Tricky Enough.