As I said in the previous post, last January I did two sessions with ChatGPT from OpenAI. In the second, my questions were directed to the following scientific topics:
• The special theory of relativity, the limit
of the speed of light in a vacuum, and the possibility of traveling in time.
Throughout the session I was asking the
following questions:
- Can
a particle travel at a speed greater than light?
Answer: According to the theory of relativity, as first described by Albert Einstein, nothing with mass can travel at the speed of light...
- What
about Cerenkov radiation?
Answer:
After explaining Cerenkov radiation, ChatGPT added this: This is different than a particle traveling faster than
the speed of light in a vacuum... The particle itself is not traveling faster
than the speed of light, but it is moving at a speed greater than the speed of
light in the medium it is travelling through, causing the Cerenkov radiation
effect. Obviously it is using the words the speed of light with two different meanings.
- What
about tachyons?
Answer:
After explaining what tachyons are, ChatGPT added that their existence is
dubious, and would raise several problems and
paradoxes… and adds: Moreover, tachyons could move into the
past, creating a paradoxical causality…
- You
said that "tachyons could move into the past". How?
Answer:
The long explanation contained this clearly wrong assertion: their time dilation (the slowing of time as perceived by
an observer moving at high speeds) would be negative.
- I
think this is wrong. Their time dilation would be imaginary.
Answer:
After apologizing for its mistake, ChatGPT explained why the Lorentz factor
would be imaginary, and added this (again wrong) assertion: time dilation as well as length contraction for a tachyon
would be described by an imaginary Lorentz factor, which would imply that time
would run backwards and distances would be imaginary for a tachyon.
- I
disagree that "time would run backwards and distances would be
imaginary for a tachyon." Both time and distances would be imaginary.
What that means for time is unknown, but time would not run backwards, as
you say.
Answer:
When I said that "time would run
backwards" for a tachyon, I was not accurate. Then it added
that imaginary time would bring problems and paradoxes.
- So
is time travel possible?
Answer:
The possibility of time travel is a topic of
much debate in physics and science fiction. The laws of physics as we currently
understand them do not rule out the possibility of time travel, but they do
place certain constraints on it. Then it talked about the
possibility of “closed timelike curves” in Relativity theory, their practical
difficulties, and possible problems posed by quantum theory.
- What
about Fermi's paradox?
Enrico Fermi
Answer:
Here came a long description of Fermi’s paradox as applied to the search for
extraterrestrial intelligences.
- I
meant Fermi's paradox applied to time travel
Answer:
I apologize for misunderstanding your
question earlier. Then came a long description of Fermi’s
paradox applied to time travel.
- What
would you say about this: quantum entanglement and Bell's theorem proves
that instantaneous transmission of information is possible? I
made this question to find out if ChatGPT would answer in the negative, as
it should.
Answer:
After a long description of quantum entanglement and Bell’s theorem, ChatGPT
wrote this: While the phenomenon of quantum
entanglement seems to imply that information can be transferred
instantaneously, it doesn't necessarily imply that information is being
transmitted faster than the speed of light… it is not yet possible to use it to
send information faster than light. I found this answer
unsatisfactory. It seems to say that it shall be done in the
future. So I formulated my next question thus:
- You
said: "it is not yet possible to use it to send information faster
than light." Will it be possible in the future?
Answer:
It is currently not possible to use quantum
entanglement to send information faster than the speed of light. The laws of
physics as we currently understand them prohibit any information from being
transmitted faster than the speed of light. But then it added
that it's always possible that in the future,
advances in our understanding of physics will allow us to harness the
properties of quantum entanglement in new and unexpected ways, but as of now
there is no theoretical or experimental evidence that it would be possible to
use quantum entanglement to send information faster than the speed of light. Therefore,
I made my last two questions:
- But
according to Relativity, it will never be possible to transmit information
at a speed greater than that of light. Right?
Answer:
ChatGPT accepted that Relativity forbids information to be sent faster than the
speed of light in a vaccuum, and added: quantum
entanglement… doesn't imply a faster-than-light communication… as of now there
is no way to use it for faster than light communication based on the current
understanding of physics.
- So
should we conclude that information will never be transmitted at a speed
greater than that of light?
Answer:
Based on our current understanding of
physics, specifically the theory of special relativity, it is believed that it
will never be possible to transmit information at a speed greater than that of
light. This is a fundamental principle of the theory of special relativity,
which has been extensively tested and is well-established by experimental
evidence.
My conclusion: It was not difficult for me
make ChatGPT fall in a few traps, and make its answers conform little by little
to what I wanted it to say. My impression, after this short test, is that this
is an excellent advance in language processing systems, with a state-of-the-art
algorithm. But any attempt to present it as something more that weak artificial intelligence is obviously
wrong. And it will need some time before it can be really useful.
ChatGPT and similar tools have been
reported as providing mistaken, incomplete or biased information.
What’s so surprising about this, as they pull their information from the
Internet, which is littered with mistaken, incomplete, or biased human-derived
information?
Some say that these tools will soon replace search engines. I hope this will not happen. When I look for information, I want to know where it comes from. Current search engines tell me that it comes from Wikipedia, or has been written by a philosophy professor, or by a Facebook user. This lets me assign a certain degree of credibility to each piece of information. I know, for instance, what to expect from Wikipedia, for I have corrected dozens of errors in this encyclopedia. If search engines were replaced by these tools, that information would no longer be available. Therefore, I would not find them useful.
The same post in Spanish
No comments:
Post a Comment