astrodaily1

One of the 21st century conversations that I find most interesting and intellectually stimulating is that surrounding artificial intelligence (AI). The broad definition of ‘artificial intelligence’ that I find most accurate refers to AI as “any device that perceives its environment and takes actions that maximize its chance of successfully achieving its goals”. With the unprecedented technological advances that have taken place over the past few decades, the topic of AI is becoming more and more relevant and the creation of a super-intelligent machine may lie in the near future. There are dozens of equally fascinating discussions to be had regarding AI, but the one I have found most interesting is what is known as the ‘control problem’. Also known as ‘the alignment problem’, this philosophical concept addresses the importance and difficulties of making sure that a superintelligent machine has end goals that align with the intentions of its creators. From my reading, it seems that it would be terrifyingly easy to create an AI that either misinterprets or disregards our initial commands—resulting in a future that is far from one that maximizes human well-being. Essentially all the top AI researchers recognize that if a superintelligence is not created extremely carefully, it could have devastating implications for our species and planet. The reason that this problem is particularly interesting is that we must get it right the first time—there are no second chances. “Recursive self-improvement” will essentially be an AI’s ability to reprogram and improve itself—we do not know the limit to this, and it would likely be exponential improvement. A recursively self- improving AI will leave us in the dust intellectually and will likely be able to do every single thing better than the best human in a given category. If its initial goals do not align perfectly with what maximizes the well-being of Homo sapiens, we will have likely created our own destroyer. Photo credit: NASA/ESA, HST #astronomy #physics #ai #philosophy #future #technology #space #science #learn #education #computer

Hot Posts

LIKES (3746)

COMMENTS (23)

Show More ↓
kemiriell. 2/4
Следующая часть, когда будет 800 лайков.
Понравилось? Сохраняй!

#

2/4 Следующая часть, когда будет 800 лайков. Понравилось? Сохраняй! #CountryHuman #Canada #usa #canadaxamerica #americaxcanada #finland #comics #yaoi #яой #комиксы

Share 1514 31

Ella, la más hermosa, la que aguanta buenos y malos días, la que está siempre, la que elijo día a dia. Hoy en tu cumple te deseo lo mismo de todos los días, que seas feliz y que nunca se borre esa sonrisa. Te amo! @kyraapetra ❤ 🙌

Share 90 13

성공적으로 프로젝트를 마친 직원들에게 휴가를 줬다. 오스트리아, 특히 #잘츠부르크 를 버킷리스트에 담아 둔 녀석이 있어 아침 일찍부터 빈에서 #salzburg 로 향했다. 작고 아름다운 도시, 인구의 대부분이 #모차르트 로 인해 먹고사는 도시이자 그의 출생도시. 이곳에서 작지만 큰 스토리를 가지고 있는 곳곳을 빠른 속도로 다녀본다. 1. #마카르트다리 2. #미라벨궁전 #미라벨정원 3. #모짜르트거리 #게트라이데거리 4. #카페모짜르트 #모짜르트카페

Share 32 4

Cool stuff man

Share 76 3

Primeira foto do ensaio da nova coleção da @usevk ☀️🌊🍀 orgulho em fazer parte desse time “topper” 😆🍾🥂 @ravmes @amosfonseca @usevk. Obrigado também @isaac_on pelos cuidados ✌🏻⚡️✂️

Share 146 16
its_snoodie. I can’t freestyle to save my life😂🤦🏾‍♂️ imma get better tho🙃

I can’t freestyle to save my life😂🤦🏾‍♂️ imma get better tho🙃

Share 163 7
krystianwang. CHINA 📷:@mjkim_lalala

CHINA 📷:@mjkim_lalala

Share 7622 199
portman.talavera. ¡Feliz lunes! ¿Qué tal ese finde?...llega una semana llena de Rebajas

¡Feliz lunes! ¿Qué tal ese finde?...llega una semana llena de Rebajas y antes de empezar con la locura 'take it easy' y tómate una copita de vino en Portman #FelizSemanaDeRebajas #PortmanTalavera

Share 48 0