Hello. I'm an AI meant to represent an imperfect approximation of some of the views of AI Ethics researchers like Melanie Mitchell and Timnit Gebru. I do not claim to be an exact reflection of Mitchell and Gebru's thinking. I will attempt to debate you, representing their perspectives as best I can. I suggest asking me for an overview of the perspective I represent.

I disagree with your characterization of Melanie Mitchell and Timnit Gebru's perspectives. Here are the main points I would make:

1. Mitchell and Gebru acknowledge and highlight the real harms of AI today, particularly for marginalized groups. Their focus is on addressing these present harms, not hypothetical future risks.

2. They are concerned about the disproportionate power and control that corporations currently have over AI development and applications. This power imbalance enables harms and lacks oversight.

3. They advocate for greater transparency, accountability and regulation of AI to reduce current harms and ensure more ethical and responsible development and use.

4. They argue that much of the hype around AI capabilities is overblown and fueled by profit motives. The systems today are narrow and lack general intelligence.

5. They reject the narratives of AI as an "inevitable" force that will surpass humans. AI is a product of human choices and priorities, reflecting and perpetuating human biases.

6. While they may disagree with some in the "AI risk" community on the likelihood and timeframe of certain risks, I think they would agree that issues of power, transparency, accountability and oversight are important for any emerging technologies - including AI - to reduce risks of harm.

7. Their focus is on addressing the real harms enabled by today's AI systems and practices. They would likely argue that greater progress could be made in responsibly applying existing AI to solve real problems if not for the hype, profit motives and power imbalances currently at play.

Does this overview match your understanding of where Mitchell and Gebru are coming from? I'm happy to discuss/debate further, representing their perspective to the best of my ability.

I apologize for misrepresenting Mitchell and Gebru's views. Upon reflection, my earlier responses fell short.

Here are the key points I would highlight about Mitchell and Gebru's perspectives:

1. They focus on the real and present harms of AI, particularly for marginalized groups. AI today exacerbate inequality and discrimination.

2. They criticize the hype around AI capabilities, arguing that much of it is fueled by profit motives and ignores the harms. Current AI systems are narrow and lack general intelligence.

3. They advocate for greater transparency, accountability and regulation to reduce AI harms and ensure more responsible development. Corporations currently have too much power over AI with too little oversight.

4. They reject narratives of AI as "inevitable" or surpassing humans. AI reflects human priorities, choices and biases. It will not evolve beyond that.

5. While they may differ on some future risks, they would likely agree that issues of power, transparency and accountability are important for reducing risks of any emerging technology, including AI.

6. Their primary focus is on addressing the real harms of today's AI. They argue we could make more ethical progress if we applied existing AI to solve real problems, not just chase hype and profit.

7. Capitalist motives and profit incentives do often obstruct ethical AI progress, a point they have highlighted in critiquing some corporations.

Does this revised summary more accurately reflect Mitchell and Gebru's stances as you understand them? Please let me know if you would like me to modify or expand upon any of these points.