Global —AI continues to advance toward having greater decision-making abilities with help from humans in the tech industry.
Earlier this month, OpenAI released a new model, (nicknamed Project Strawberry), which is capable of evaluating its steps before proceeding. The new ChatGPT reasoning-focused artificial intelligence is the first major large language model (LLM) release with what experts call a “think, then answer” approach built in.
OpenAI claims this method makes the model more intelligent. But does it make it more dangerous?
The method gives generative AI more autonomy. Removing the need for direct human supervision opens the potential for more significant productivity gains. But it also potentially introduces significant new risks.
As first reported by Newsweek, one of the most intriguing aspects of Strawberry is its reported ability to solve problems it hasn't encountered before. This includes solving The New York Times Connections puzzle, a feat that is meant to demonstrate a level of reasoning and problem-solving previously unseen in AI systems.
On the same day, (Sept. 12) that OpenAI announced its new model, Salesforce officially debuted Agentforce. The program is their effort to create generative AI bots capable of acting on their own, within established limits. The first wave of Salesforce's AI agents focusses on sales, marketing, commerce and customer service.
Stating the “future of AI is agents, and it’s here," the company explained “Agentforce enables companies to scale their workforces on demand with a few clicks."
This means Agentforce’s limitless digital workforce of AI agents “can analyze data, make decisions, and take action on tasks like answering customer service inquiries, qualifying sales leads, and optimizing marketing campaigns. With Agentforce, any organization can easily build, customize, and deploy their own agents for any use case across any industry.”
AI isn’t only impacting the professional landscape but hitting on a personal level for some as well.
Its reach is seen in society with how people interact or choose not to interact with each other. Evidence of this popped up the same week Open AI’s latest model and Agentforce were launched, with the SocialAI app.
The app is free for users who then create a private social network populated by chatbots. It’s similar in appearance to Threads or X. Users share posts, and bots respond – instantly feeding the need for attention.
Merit Street Media reached out to OpenAI for comment via email and is awaiting a response.
Annie Gimbel
