A top-ranked executive at OpenAI admitted the need to involve government regulators when designing and implementing oversight over artificial intelligence projects. Mira Murati, who currently serves as the chief tech officer (CTO) within OpenAI, challenged the AI developers to accommodate the input of government regulators.
Necessity to Include Government Regulators in Formulating Standards
The CTO confessed that OpenAI should consider beyond the pause letter petitioned against the firm’s GPT-4 model. The admission is timely as OpenAI gears towards artificial general intelligence.
Murati considers it critical for government regulators to consider their detailed involvement in formulating safety standards. Such is necessary to involve regulators in formulating standards that would guide the deployment of the advanced ChatGPT model.
Murati delves into the letter cosigned by Twitter Inc.’s chief executive, alongside 2600 tech leaders and researchers demanding a six-month pause for further release of advanced AI. The tech expert commented during the April 24 interview facilitated by the Associated Press.
Building Safer Artificial Intelligence Systems
The chief tech official at OpenAI considers that building safer systems is unattainable through the closure of AI development. She considers such to represent a hypothetical intellectual standard involving the artificial agent in executing tasks that involve intelligence blended with human-level cognition.
The Associated Press interviewer tasked Murati to illustrate her perspective regarding the safety precautions undertaken by OpenAI before the GPT-4. She explained that the OpenAI undertook a slow training approach to avoid inhibiting the machine’s undesirable behavior.
Murati revealed that the slow training process would facilitate the discovery of downstream concerns linked with such changes. She noted the need to exercise caution to avoid creating imbalances, hence the need to integrate constant audits. It implies that the training is conducted cautiously with prompt assessment in every intervention.
Is Safety Guaranteed After Six-Month Pause on AI Development?
Murati acknowledged the stance adopted by the tech executives expressing uncertainty of the unknown, particularly on future AI impact. The CTO observed that such individuals would lobby for increased government regulation or the Musk-led resolution for a six-month pause on further AI development.
The uncertainty surrounding the impact of AI on social order is garnering attention from researchers and tech executives led by Elon Musk, Eliezer Yudkowski, and Steve Wozniak. The letter signed by 2600 executives has since turned divisive, with Bill Gates and Andrew Ng condemning the opposition toward AI.
Murati echoes the Musk-led faction petitioning for the government’s involvement in regulating AI systems. She adds that OpenAI continually engages governments and regulators to agree on certain standards.
Murati portrayed defensive in his responses, particularly criticizing the developmental pause. She dismissed claims that OpenAI was training a GPT-5 model. Terming the statement false, the tech executive revealed that OpenAI did not harbor such plans for the next six months. She ruled out rushing the GPT4 model by stating that it took six months to focus solely on safe development.
GPT-4 Model Yet to Border the General AI
Murati emphasizes that artificial intelligence and current models such as GPT4 are yet to realize the desired reliability and safety levels. The statement dismissed the likelihood of the GPT-4 model bordering on general intelligence. Safety concerns and delay in training GPT-5 illustrated that the discovery of general intelligence is miles away from the accomplishment.
Murati’s statements urging the involvement of regulators in formulating standards to guide the development and deployment of AI coincide with the increased scrutiny by various European governments. Recently, German and Spanish authorities joined Italy in opening a probe into GPT products and OpenAI compliance with data privacy laws.
OpenAI faces a bleak future, considering that Italy already issued an April 30 deadline for the AI firm to comply with the EU General Data Protection Regulation (GDPR).
The imposition of the bans could impact the region’s crypto scene, given the advanced adoption of trading bots established upon applications that leverage GPT API. The increased government scrutiny of OpenAI could prompt a mass exodus of similar companies outside Europe.
Editorial credit: Ascannio / Shutterstock.com