OpenAI is extending its Deep Research feature to ChatGPT Plus, Team, Education, and Enterprise users, offering 10 queries per month initially to these users.
Deep Research uses OpenAI's specialized o3 model to assist with complex research tasks by analyzing text, images, and PDFs to produce comprehensive reports.
With the emergence of China's DeepSeek and the open-sourcing of its DeepSeek-R1 model under an MIT license, competition in the AI landscape has intensified.
While DeepSeek adopts an open-source approach, OpenAI continues to offer powerful capabilities through subscription tiers, prompting a shift in the market.
Different companies like Perplexity, Anthropic, and OpenAI offer unique AI-powered research solutions, each with its own approach and features.
The integration of DeepSeek-R1 by Perplexity at a lower price point than OpenAI showcases the competition and the potential of open approaches in AI development.
Deep Research challenges conventional AI models by offering significant advancements, with a 26.6% accuracy on the challenging 'Humanity’s Last Exam' benchmark.
Despite its advanced capabilities, Deep Research has limitations such as consensus bias and reliance on existing web content, potentially hindering its analysis in certain domains.
Organizations integrating Deep Research need to redefine their information workflows and view it as an amplification of human capabilities rather than a replacement for human judgment.
The pricing structure of Deep Research, with varying query limits based on subscription tiers, forces organizations to prioritize questions and strategize its application.
The democratization of deep research capabilities will shift competitive advantage from information access to how organizations utilize AI-generated insights in their decision-making processes.