The news: DeepMind CEO Demis Hassabis said a private beta version of its chatbot Sparrow could be released in 2023, per The Independent.
- In September, the Alphabet subsidiary postponed Sparrow’s release over public safety concerns.
- DeepMind says Sparrow has features that ChatGPT lacks, including an ability to cite sources through reinforcement learning.
The announcement coincides with Microsoft’s deepening partnership with OpenAI to integrate ChatGPT into all of its tools.
- On Monday, Microsoft said that it will allow more customers to access OpenAI’s tools through the cloud in its Azure OpenAI Service, including potentially ChatGPT Professional via an upcoming trial release, per Insider.
- Companies like KPMG, Al Jazeera, and Moveworks are already using the service.
Much grander ambitions: DeepMind and OpenAI aren’t merely working on trendy chatbots. Both companies have their eyes fixed on achieving artificial general intelligence (AGI).
- Part of the impetus for Microsoft’s investment in OpenAI is to build AGI.
- “The creation of AGI will be the most important technological development in human history, with the potential to shape the trajectory of humanity,” OpenAI CEO Sam Altman said in 2019, per Acceleration Economy.
A dystopian future? Both Hassabis and Altman have issued stark warnings alongside their grand plans.
- Altman said AI’s ability to generate human-like responses could be abused, warning of “scary moments” ahead and “significant disruptions” to society, per The Independent.
- Meanwhile, Hassabis told Time that AI is “on the cusp” of significantly damaging humanity.
- With an education system reeling from ChatGPT’s release, AI is already shaping humanity’s trajectory.
- AI companies’ acknowledgment of danger is insufficient if not accompanied by more safeguards for the technology.
- In the absence of precautionary measures, expect more lawsuits over generative AI.