Do They Exist?

Extraterrestrial life and artificial intelligence have long been staples of science fiction. But is there any reason to think we might someday encounter such beings in reality? Read on to learn about the emerging fields of AI research and astrobiology, and discover why science fiction may soon become science fact.

AI: Can machines have minds?

“Artificial Intelligence #1″ by Francesco Romoli / cropped from original
“Artificial Intelligence #1,″ © Francesco Romoli / cropped from original

Thanks to incredible advances in computer science over the last several decades, machines and programs capable of simple cognitive tasks are now commonplace in daily life. However, such devices generally work by using preprogrammed knowledge or sets of rules to simulate intelligence, rather than by actually thinking on their own. For this reason, they are sometimes described as “weak AI.”

"The original question, 'Can machines think?' I believe to be too meaningless to deserve discussion. Nevertheless I believe that at the end of the century... one will be able to speak of machines thinking without expecting to be contradicted." (Alan Turing, "Computing Machinery and Intelligence" (1950) - photo credit: <a href="">Sjoerd Ferwerda</a>, used under <a href="">CC BY-SA 3.0</a>) 391w" sizes="(max-width: 133px) 100vw, 133px" /> Sculpture of Alan Turing, who proposed a test to measure a computer’s intelligence

“Strong AI,” machines with minds of their own, capable of independent thought and learning, are another matter. Simply defining a “mind,” let alone detecting or creating one, remains a highly elusive task for scientists. Nonetheless, there are many computer scientists who believe that strong AI is possible, and have devoted entire careers to producing it.

One of the first to seriously consider strong AI was Alan Turing (1912-1954), one of the founding fathers of computer science. Turing did not believe we could ever truly know whether AIs had minds, but thought they could be programmed to behave as if they did. He devised a test to measure a machine’s intelligence, based on whether or not it could pass for human. To pass the Turing Test, an AI must successfully convince a human interviewer, through typed conversation, that it too is human.

See if you can tell a human from an AI! Introduce yourself in the textbox below. Someone will respond shortly. Your job is to determine, through conversation, what type of being they are.

Which did you just talk to?


Click image to visit Cleverbot's homepage

Correct! You just spoke with Cleverbot, one of the world's most advanced AIs. Though you weren't fooled, many have been. At the 2011 Techniche conference in India, a more powerful, offline version of Cleverbot managed to convince interviewers it was human 59.3% of the time! For comparison, human participants only passed 63.3% of the time. In other words, Cleverbot was almost as good at passing for human as humans themselves!

URL used with permission from Rollo Carpenter, creator of Cleverbot. Please do not copy. To embed Cleverbot on your own website, use one of the links at


Click image to visit Cleverbot's homepage

Nope! You've been fooled by Cleverbot, one of the world's most advanced AIs. Don't feel bad though, you're not the only one. At the 2011 Techniche conference in India, a more powerful, offline version of Cleverbot managed to convince interviewers it was human 59.3% of the time! For comparison, human participants only passed 63.3% of the time. In other words, Cleverbot was almost as good at passing for human as humans themselves!

URL used with permission from Rollo Carpenter, creator of Cleverbot. Please do not copy. To embed Cleverbot on your own website, use one of the links at

Does Cleverbot’s ability to pass the Turing test with such high accuracy prove that it can think? Not necessarily. Cleverbot works by comparing what users say to a database of millions of lines of text, from which it selects the most relevant response. With a database large enough to contain every possible response to every possible question, a machine like Cleverbot could hold its own in any conversation without thinking at all about what it says. In all fairness to Cleverbot, it also stores information from each new conversation, honing its ability to say the right thing in any given context. In other words, Cleverbot can learn! Simple learning AIs are now common in online settings, such as websites that recommend content based on users’ browsing history.

The phrase, "to pull oneself up by one's bootstraps," means to better oneself by individual effort. In the context of AI, such self-betterment includes the ability to gain knowledge and cognitive skills, perhaps including consciousness. Of course, to literally pull oneself up by one's bootstraps is physically impossible. Does this imply strong AI might be an impossible dream? (Photo credit: <a href="">jimshooz7</a>, used under <a href="">CC BY 2.0</a>) 640w" sizes="(max-width: 230px) 100vw, 230px" /> “Bootstrapping” comes from the phrase, “pull yourself up by your bootstraps.”

Some more advanced programs can learn not just specific phrases, but generalized patterns and rules as well. AIs with this ability can gradually improve themselves, acquiring new skills and gaining intelligence over time, a process known as bootstrapping. In theory, bootstrapping could allow an AI to develop a full independent mind on its own, long before we understand how to program it ourselves!

To accelerate this process, some scientists propose using a form of natural selection. By simulating challenges like those faced by our own ancestors, they could encourage learning machines to adapt in similar ways. Eventually, they hope this would lead AIs to evolve minds. Although biological evolution takes millions of years, existing supercomputers can simulate thousands of “generations” in a matter of minutes. The rise of strong AI may be just over the horizon!

“Weak” AIs that can perform simple cognitive tasks are now commonplace. Some can even pass for human in conversation. Yet “strong” AIs with minds of their own do not yet exist. Some scientists believe AIs could be programmed to evolve minds themselves through “bootstrapping.”


Gallery: explore the past, present, and future of AI research!

Previous Image
Next Image

info heading

info content

ET: Is there anybody out there?

(Left: Arecibo Observatory in Puerto Rico, a radio telescope which collects data for SETI, the Search for Extraterrestrial Intelligence. Right: “Exoplanetscape 14,” © John P. Alexander, imagining a sentient alien life-form based on superconducting crystals, on the cold moon of a distant exoplanet.)

Astrobiology, the study of life’s place in the universe and possible existence beyond Earth, has been a subject of scientific interest since the beginnings of space exploration. In 1960, astronomer Frank Drake (born 1930) became one of the first to actively search for extraterrestrial life. Using radio telescopes, he listened for signals broadcast by alien civilizations. He also created an equation to calculate the number of such civilizations in our galaxy: In this version of the Drake Equation, n* is the number of stars that live long enough for life to evolve, fp is the fraction of stars with planets, ne is the number of worlds per star that can support life, fl is the fraction of worlds that actually do support life, fi is the fraction of inhabited worlds with intelligent life, fc is the fraction of intelligent organisms to develop radio technology, and L is the average lifetime of such a civilization divided by the age of the universe. 300w, 1024w" sizes="(max-width: 1069px) 100vw, 1069px" />The equation cannot be definitively solved, since many of the factors remain unknown. Nonetheless, it can be used to estimate the number of alien civilizations that might exist, based on what we do know about our own civilization, life on Earth, and other planets and stars. With the calculator below, you can use the Drake Equation to estimate for yourself how many alien civilizations exist in the Milky Way!

By the most hopeful estimates of the Drake Equation, there could be many thousands or even millions of other civilizations in our galaxy alone! Yet we have been listening for them since 1960, and so far haven’t heard anything. Where are they? This problem is known as the Fermi Paradox, and has several possible answers:

  • Technologically advanced civilizations aren’t very common after all
  • Civilizations don’t usually last very long, collapsing within a short period of time
  • Alien civilizations capable of broadcasting signals choose not to, perhaps for fear of attracting dangerous invaders
  • Alien civilizations do communicate, but with technologies other than radio signals we can detect

Listening for radio signals is not the only way astrobiologists search for life elsewhere in the universe. Some look within our own solar system, aided by mechanical probes sent to other planets and moons. Others study exoplanets (planets outside our solar system), of which nearly 2,000 are now known. Using spectroscopy, the analysis of starlight passing through their atmospheres, astrobiologists may soon be able to search for biosignatures, evidence for gases similar to those produced by life on Earth.

SETI, the Search for Extraterrestrial Intelligence, listens for radio signals coming from alien civilizations. Astrobiologists also search for life within our own solar system, and for signs of life in the atmospheres of extrasolar planets.


Photo gallery: the search for life, in our solar system and beyond!
(Note: if first slide does not display caption, or appears with large blank space underneath, scroll down and skip to the second, then return to the first. It should now display properly.)

Previous Image
Next Image

info heading

info content

Will we ever meet them?

Using spectroscopy, astrobiologists hope to discover alien life in nearby solar systems within just a few decades! We won’t know much else about it, though, except that it’s enough like Earth life to alter its atmosphere in recognizable ways. If SETI should detect a radio signal from an extraterrestrial civilization, we might be able to learn much more, and perhaps even send a reply, if we can successfully decode it! Direct contact with alien life is another matter, however. The nearest star, Proxima Centauri, is just over four light years away, meaning that even at light speed it would take four years to travel there. The fastest spacecraft ever launched would take 80,000 years to reach it! Of course it’s always possible that alien explorers could come to us, perhaps drawn by our own radio signals. But otherwise, our only hope for direct contact in the foreseeable future is probably to discover alien life within our own solar system.

Between 1958 and 1964, physicists at Project Orion designed several starships that, using nuclear explosions to accelerate up to 10% the speed of light, could get to Proxima Centauri in as little as 44 years! Such starships would be enormous, expensive, and potentially dangerous to launch, and of course sustaining human life in space for 44 years would be another challenge altogether. However, they could potentially have been built with existing technology, had Project Orion not been scrapped following the Partial Test Ban Treaty of 1963. (Image from Ascension, © 2014 SyFy Media LLC, <a href="">fair use</a>)
A nuclear-powered starship in the miniseries Ascension
Though it may sound like a fantasy out of Peter Pan, it may in fact be possible to sail to the nearest stars! Though we don't notice it, light produces a small amount of physical force as it reflects off a surface, known as "photon pressure." In 2010, the Japan Aerospace Exploration Agency successfully used this pressure to power the first "solar sail" spacecraft. Known as IKAROS, it uses a vast but lightweight sail to collect photon pressure, which pushes it through space. In 2011, NASA successfully launched a solar sail of its own, the NanoSail-D2. (Photo credit: <a href="">JAXA</a>) 705w" sizes="(max-width: 182px) 100vw, 182px" /> IKAROS, the first spacecraft to use light sails, in a “selfie” taken by ejecting a camera

Does this mean humans will never be able to travel quickly to other stars? No! Designs for nuclear-powered starships, which could travel up to 10% the speed of light, have existed since the 1960s. Had it been practically possible to build and launch them, we could already have humans or robots in the nearest star system today! A new private space exploration project called Breakthrough Starshot, announced by entrepreneur Yuri Milner in April 2016, aims to launch thousands of small probes toward Proxima Centauri within 20 years. These probes are to use “light sails” propelled by lasers, accelerating them up to 20% the speed of light! Theoretically, light sails could someday be used for larger, manned starships as well. Some physicists have even proposed “faster than light” methods of space travel: “warp drives” that would rapidly distort spacetime around a starship, or wormholes naturally leading to distant parts of the universe.

Nuclear or light-powered starships would be extremely challenging to build, and it isn’t yet known if warp drives or wormholes are possible at all. But if projects like Breakthrough Starshot are successful, some of us may live to see the first robotic exploration of other star systems… and, if manned starships ever prove feasible, perhaps our grandchildren will get to visit them in person someday!

If life exists around nearby stars, we may know within the next twenty years. Direct contact with alien life seems much less likely. However, this could change if we find life in our own solar system, aliens come to us, or if we develop technology to travel quickly to other stars.

See References


Among Us Already…

Print Friendly, PDF & Email