# Better Datasets May Accelerate AI by an Order of Magnitude by Alexander Wissner-Gross > “Most importantly, the prioritized cultivation of high-quality training datasets might allow an order-of-magnitude speedup in AI breakthroughs over purely algorithmic advances. For example, we might already possess the algorithms and hardware that will enable machines in a few years to author human-level long-form creative compositions, complete standardized human examinations, or even pass the Turing Test, if only we trained them with the right writing, examination, and conversational datasets. Additionally, the nascent problem of ensuring AI friendliness might be addressed by focusing on dataset rather than algorithmic friendliness—a potentially simpler approach.” > **— Alexander Wissner-Gross**, *2016, Edge Annual Question, “What Do You Consider The Most Interesting Recent [Scientific] News? What Makes It Important?”* ## Sources and Context - **Direct Edge response and surrounding essay:** [Datasets Over Algorithms](https://www.edge.org/response-detail/26587) — Edge published this exact passage in its 2016 Annual Question collection, *WHAT DO YOU CONSIDER THE MOST INTERESTING RECENT [SCIENTIFIC] NEWS? WHAT MAKES IT IMPORTANT?*. The response page preserves the surrounding argument and qualifications. - **Complete local collection index:** [[research/Edge 2016 Full Quotation Curation|Edge 2016 Full Quotation Curation]] — retained source-first record for the complete response cohort and this passage's promotion status. ## Related Articles and Collections - **Collection:** [[collections/Edge|Edge]] - **Collection:** [[collections/Machine Succession|Machine Succession]] - **Article:** [[articles/Modern Artificial Intelligence in the 1970s|AI in the 1970s]] - **Article:** [[articles/Christopher Nolan's Odyssey at the Threshold of Machine Succession|Nolan's Odyssey]] - **Article:** [[articles/Cyber-Physical Lessons from Imperial Beekeeping|Cyber-Physical Beekeeping]] - **Wiki map:** [[wiki/Artificial Intelligence|Artificial Intelligence]] ## Related Topics - [[wiki/Artificial Intelligence|Artificial Intelligence]] - [[wiki/Infrastructural Superintelligence|Infrastructural Superintelligence]] - [[wiki/Machine Learning|Machine Learning]] - [[wiki/Machine Succession|Machine Succession]] ## Share on Social Media ``` “Most importantly, the prioritized cultivation of high-quality training datasets might allow an order-of-magnitude speedup in AI breakthroughs over purely algorithmic advances. For example, we might already possess the algorithms and hardware that will enable machines in a few years to author human-level long-form creative compositions, complete standardized human examinations, or even pass the Turing Test, if only we trained them with the right writing, examination, and conversational datasets. Additionally, the nascent problem of ensuring AI friendliness might be addressed by focusing on dataset rather than algorithmic friendliness—a potentially simpler approach.” — Alexander Wissner-Gross, 2016, Edge Annual Question, “What Do You Consider The Most Interesting Recent [Scientific] News? What Makes It Important?” https://bryantmcgill.com/simple-reminders-alexander-gross-better-datasets-accelerate-ai-order ```