Microsoft, OpenAI and more come together for election security

Microsoft, OpenAI, Google and more are coming together for election security. The focus here is artificial intelligence-generated fake content. will be.

Microsoft, Meta, Google, OpenAI, Amazon, Adobe, IBM, ARM, Stability AI, TikTok, X, Snap, LinkedIn and seven other AI and technology companies, “AI Selection AgreementThey came together around. Companies that signed this agreement, developing methods to detect deepfakeswhich may lead to deepfakes identifying modelsdeepfakes aimed at manipulating the election prevent its spread on social platforms and is clearly committed to being transparent with the public regarding false/fake election information. This is really important because during the election campaign, there are many politicians who try to put the other side behind with fake content.


Behind ChatGPT OpenAI He recently made a splash with his statement, “GPT-4 offers a little more than the internet in the production of biological weapons.” OpenAI, What else will it offer commercially other than ChatGPT, which currently has many competitors? The news here According to the company, productive artificial intelligence It is working to make it a system that can handle complex tasks automatically..

There are still no excessive details, but it is stated that the company has developed two different artificial intelligence agents (systems that perceive and can react) within the scope of these new plans. It is reported that the first artificial intelligence agent will be used locally on a computer to automate complex tasks. For example, it is mentioned that the system can take data from a document and transfer it to a table or fill out expense reports and enter them into accounting software.

By providing special permissions here It is said that the entire system/computer control will be handed over to the artificial intelligence agent. The second artificial intelligence agent is expected to collect public data, create routes, or book flight tickets. will focus directly on performing web-based tasks It is said.

It is unknown when these studies will be commercialized. OpenAI, He recently came to the fore with his statement that “GPT-4 offers a little more than the internet in the production of biological weapons.” OpenAI made important statements within the scope of an internal research.

Accordingly, the big language model named “GPT-4”, which is the basis for the paid ChatGPT developed by the company, It gives people only a slight advantage over the information available on the internet when it comes to researching and producing biological weapons. In other words, when it comes to producing biological weapons for people, ChatGPT does not provide too much detail compared to Google results, but information can still be obtained.

Stating that there should be no worry about this issue, but stating that the protection systems in the process will be improved. because the company still poses some vital risksas experts often say, “GPT-4 could significantly aid terrorists, criminals and other miscreants” does not share his concern, but the situation does not seem to be exactly like that in reality.