Bloomberg adds ChatGPT-inspired software to its terminals, while Amazon invests in generative AI during a major company restructure.
In The Terminal, there’s a ghost in the machine—or is it a robot? A Bloomberg representative told CNBC on Thursday that the business intends to incorporate ChatGPT-inspired software into its Bloomberg terminal offering. The same day, Amazon CEO Andy Jassy declared the company is “investing heavily” in generative AI while cutting 27,000 positions and other elements of its business.
Rather than depending on OpenAI, the organization that created ChatGPT, Bloomberg has created its own software. This indicates that the generative AI tool has been trained using a smaller sample of data than ChatGPT. The goal of this training is to improve the financial knowledge of the tool. AI tools have a history of being imprecise when it comes to discussing financial topics or even basic mathematics. Therefore, using a smaller dataset for training could be a wise decision.
Given Amazon’s everything, everywhere, all at once business model, it is not surprise that the company has used generative AI more broadly:
- Jassy suggested incorporating generative AI into the user experience, but she did not specify how. We recommend utilizing AI to stop Amazon from suggesting things you’ve bought.
- Even if its generative AI product disappoints consumers, Amazon is retargeting the same gold rush businesses. Amazon’s cloud computing business, AWS, has released tools to encourage developers to use its services to build massive language models and generative AI tools.
Jassy looks to be following AWS’s lead in supporting internet infrastructure by running computationally intensive applications like ChatGPT.
There are statistics suggesting that ICT contributes more to climate change than aviation globally. AI expert Professor Sandra Wachter shared this with The Daily Upside. Professor Wachter also noted that the energy required for AI has increased an estimated 300,000 times between 2012 and 2018. In addition, there are other numbers indicating the significant amount of energy consumption required for AI. For example, a single training session for ChatGPT, the AI model that I am based on, requires the same amount of energy consumption as 126 Danish homes in a year.
As competitors follow OpenAI’s lead, regulators are cracking down. Italy stopped using ChatGPT last month due to privacy concerns. They provided the company with a list of requirements to meet by April 30 to avoid further bans. The European Union’s highest data protection entity formed a task force to monitor ChatGPT. Meanwhile, the French privacy regulator is reviewing complaints related to the company.