filmov
tv
Ep. 21 | Synthetic data, privacy, and the future of trust
Показать описание
Over the last decade, the business of data has disrupted nearly every business category with its promise of technological, industrial, and human advancement. Data continues to captivate our interest as entrepreneurs, executives, and policymakers for its potential to democratize the next wave of productivity with artificial intelligence and machine-to-machine advancements. To advance this wave of productivity, new models of data have been invented: Synthetic Data.
As it suggests, synthetic data is completely artificial and offers the promise of both usefulness and privacy. Artificial intelligence that is trained on real-life information often contains a baked-in bias: algorithmic decision-making in fields such as criminal justice and credit scoring shows evidence of racial discrimination. The promise of synthetic data allows organizations and governments to overcome geographical, resource, and political barriers. It can be applied to solving some of the world’s biggest problems, from international medical research, fairness in lending, to reducing fraud and money laundering. By 2022, Gartner estimates over 25% of training data for AI will be synthetically generated. It is already being used in healthcare, banking, crime detection, manufacturing, telecom, retail, and several other fast-moving industries to accelerate learning.
However, its usefulness hinges on privacy: that anybody utilizing synthetic data could make the same statistical decisions as they would from the true data — without being able to identify individual contributions.
On this episode of the GeoTech Hour, experts will discuss how if the privacy thresholds can be legally and ethically addressed, synthetic data can be the best way to safely unlock the potential of the data economy.
--------------------------------------------------------------------------------------------
Driven by our mission of “shaping the global future together,” the Atlantic Council is a nonpartisan organization that galvanizes US leadership and engagement in the world, in partnership with allies and partners, to shape solutions to global challenges.
As it suggests, synthetic data is completely artificial and offers the promise of both usefulness and privacy. Artificial intelligence that is trained on real-life information often contains a baked-in bias: algorithmic decision-making in fields such as criminal justice and credit scoring shows evidence of racial discrimination. The promise of synthetic data allows organizations and governments to overcome geographical, resource, and political barriers. It can be applied to solving some of the world’s biggest problems, from international medical research, fairness in lending, to reducing fraud and money laundering. By 2022, Gartner estimates over 25% of training data for AI will be synthetically generated. It is already being used in healthcare, banking, crime detection, manufacturing, telecom, retail, and several other fast-moving industries to accelerate learning.
However, its usefulness hinges on privacy: that anybody utilizing synthetic data could make the same statistical decisions as they would from the true data — without being able to identify individual contributions.
On this episode of the GeoTech Hour, experts will discuss how if the privacy thresholds can be legally and ethically addressed, synthetic data can be the best way to safely unlock the potential of the data economy.
--------------------------------------------------------------------------------------------
Driven by our mission of “shaping the global future together,” the Atlantic Council is a nonpartisan organization that galvanizes US leadership and engagement in the world, in partnership with allies and partners, to shape solutions to global challenges.
0:56:19
0:46:49
0:10:24
0:00:21
0:24:54
0:01:56
0:05:33
0:02:25
0:25:49
0:04:48
0:14:24
0:00:39
0:00:57
0:06:18
0:00:59
0:55:08
0:26:24
0:00:59
0:00:56
0:35:36
0:01:58
0:27:03
0:12:48
1:02:08