The rise of AI has really instilled plenty of questions in people’s minds. And for starters, we’re talking about how it just might cause the downfall of human competency, thanks to its great capabilities.
Many AI systems are being called out as super powerful and they’re posing significant risks to all of humanity. And if AI’s goals continue to behave in this manner, the end might be near, many fear.
To avoid such ordeals and us getting close to doomsday, we’ve got the co-founder of DeepMind explaining how there is a solution possible. And that entails ruling out several capabilities linked to the world of AI.
Thanks to a recently published interview, Mustafa Suleyman explained how integral it has become to rule out some aspects linked to self-improvement. And that’s related to the chance of AI to get better with time.
You would never wish to have the AI go off in different directions and enable code updates without anyone providing oversight. Moreover, it might be a licensed activity and it’s quite like handling anthrax.
There has been a lot of focus on this subject and how AI regulation is taking place at an institutional level. Moreover, both Musk and Zuckerberg are gathering together in the capital city of Washington to take part in a new AI forum that’s arising behind closed doors.
The main problem is people aren’t setting boundaries, he added and there needs to be some limit in place that the world of AI shouldn’t cross. These types of boundaries establish provable safety including the actual code, all the way down to interacting with so many others or even with the likes of humans. This entails all sorts of motivations as well as incentives having to do with firms establishing such technology.
In the past, we saw Suleyman being the co-founder of Inflection AI. This brings forward a new chatbot that’s programmed to put out emotional support. And yes, it might not be as fabulous as other places, but this chatbot is said to be highly controllable.
Remember, you need to be super optimistic about AI and ensure it’s regulated effectively. And thanks to the encouraging words of Suleyman, he seems like he’s more than brave about what’s taking place.
Speaking to the publication, there are close to 101 issues and those need more focus and attention than just this. Be it privacy or facial recognition that’s overshadowed by bias, or maybe even online moderation, there’s a lot worth considering.
Moreover, Suleyman seems to be just one of the many experts in this field who sounds a little off in terms of regulating AI. One other cofounder of DeepMind explained how AI technology development needs to be done in a careful manner through such a method and entails plenty of experiments as well as testing.
Meanwhile, the CEO of Microsoft added that the manner to avoid this kind of Runaway AI is to ensure we begin using categories where humans are at the forefront.
Remember, since the start of March of this year, we had nearly 34,000 individuals, including godfathers of AI sign open letters hailing from Future of Life Institutes that are non-profit. They call upon AI labs to put a pause on training of any kind of technology which is more competent than OpenAI which is rolling out GPT-4.
So to sum it up in a nutshell, we can be hopeful that AI is not putting an end to the human race, provided the right kind of regulation is in place.
H/T: Yuval Noah Harari / The Economist. / YT
Read next: Meta's Encryption Plans: UK's Safety Concerns vs. Tech Titans' Privacy Push
Many AI systems are being called out as super powerful and they’re posing significant risks to all of humanity. And if AI’s goals continue to behave in this manner, the end might be near, many fear.
To avoid such ordeals and us getting close to doomsday, we’ve got the co-founder of DeepMind explaining how there is a solution possible. And that entails ruling out several capabilities linked to the world of AI.
Thanks to a recently published interview, Mustafa Suleyman explained how integral it has become to rule out some aspects linked to self-improvement. And that’s related to the chance of AI to get better with time.
You would never wish to have the AI go off in different directions and enable code updates without anyone providing oversight. Moreover, it might be a licensed activity and it’s quite like handling anthrax.
There has been a lot of focus on this subject and how AI regulation is taking place at an institutional level. Moreover, both Musk and Zuckerberg are gathering together in the capital city of Washington to take part in a new AI forum that’s arising behind closed doors.
The main problem is people aren’t setting boundaries, he added and there needs to be some limit in place that the world of AI shouldn’t cross. These types of boundaries establish provable safety including the actual code, all the way down to interacting with so many others or even with the likes of humans. This entails all sorts of motivations as well as incentives having to do with firms establishing such technology.
In the past, we saw Suleyman being the co-founder of Inflection AI. This brings forward a new chatbot that’s programmed to put out emotional support. And yes, it might not be as fabulous as other places, but this chatbot is said to be highly controllable.
Remember, you need to be super optimistic about AI and ensure it’s regulated effectively. And thanks to the encouraging words of Suleyman, he seems like he’s more than brave about what’s taking place.
Speaking to the publication, there are close to 101 issues and those need more focus and attention than just this. Be it privacy or facial recognition that’s overshadowed by bias, or maybe even online moderation, there’s a lot worth considering.
Moreover, Suleyman seems to be just one of the many experts in this field who sounds a little off in terms of regulating AI. One other cofounder of DeepMind explained how AI technology development needs to be done in a careful manner through such a method and entails plenty of experiments as well as testing.
Meanwhile, the CEO of Microsoft added that the manner to avoid this kind of Runaway AI is to ensure we begin using categories where humans are at the forefront.
Remember, since the start of March of this year, we had nearly 34,000 individuals, including godfathers of AI sign open letters hailing from Future of Life Institutes that are non-profit. They call upon AI labs to put a pause on training of any kind of technology which is more competent than OpenAI which is rolling out GPT-4.
So to sum it up in a nutshell, we can be hopeful that AI is not putting an end to the human race, provided the right kind of regulation is in place.
H/T: Yuval Noah Harari / The Economist. / YT
Read next: Meta's Encryption Plans: UK's Safety Concerns vs. Tech Titans' Privacy Push
