Up next in 10
Three physically allowed endings for the universe: heat death, a Big Rip, and a Big Crunch. One number, w, picks among them.
On August 26, 2026, a Cornell-led model put a crunch back on the clock: expansion stops in about 11 billion years. That is a model, not a verdict. This video walks the paperwork, not the slogan.
0:00 The clock at zero
0:50 The Number That Decides
1:50 How Space Actually Stretches
3:00 The Freeze Path
4:15 Last Light
5:30 Phantom Energy
6:45 The Part That Actually Matters
8:00 The Crunch Comes Back
9:30 What Does Not Happen
10:30 Which Clock You Are On
This is a cosmology explainer, not a doomsday forecast. None of these endings is a next-decade event. Figures are from public papers, including Planck 2018, Caldwell, Kamionkowski, and Weinberg 2003, Adams and Laughlin 1997, DESI 2024, and Tye and collaborators 2025. Estimates are marked as about.
Tags: cosmology, dark energy, heat death, Big Rip, Big Crunch, DESI, Planck, universe ending
#cosmology #darkenergy #universe #bigrip #heatdeath #bigcrunch #science #DESI
Disclaimer: Educational science content about far-future cosmological models. Not a prediction of near-term catastrophe.
Visuals are original AI-generated stills. Audio may be synthetic. Not medical, legal, or financial advice.
Show More Show Less View Video Transcript
0:00
On August 26, 2026, a Cornoled team put
0:04
a clock on an ending most textbooks had
0:06
written off. Their model fitted to
0:08
recent maps of dark energy said the
0:10
expansion of space stops in about 11
0:13
billion years. Then the universe falls
0:15
back in. Total age at the crunch about
0:18
33 billion years. We have already spent
0:21
13.8 billion of those. That is not a
0:24
forecast that the sky is collapsing this
0:26
century. It is one of three endings
0:28
physics still allows and they all hang
0:30
on a single parameter called W. At T
0:33
equals 0, you are standing on Earth. The
0:36
microwave left over from the Big Bang is
0:38
2.725 Kelvin. Empty space is stretching
0:42
at about 70 km/s for every mega parseek
0:45
of distance. At t= 1 second, nothing you
0:49
can feel has changed. At t= 5 seconds,
0:52
still nothing. The universe does not end
0:55
on a kitchen timer. It ends on that
0:57
number. The number that decides W is the
1:00
ratio of pressure to energy density in
1:03
whatever is pushing space apart. A
1:05
cosmological constant sits at minus1.
1:08
Ordinary matter sits near zero.
1:10
Radiation sits at plus 1/3. Cross below
1:14
minus1 and you enter phantom energy, a
1:17
density that grows as space grows. Climb
1:20
above minus one/3 long enough and
1:22
gravity can win the tugofwar. Plank's
1:25
2018 cosmological fit. Put W at about
1:28
minus 1.03. Plus or minus 0.03.
1:33
That error bar is the whole argument.
1:35
Hold at minus1 and you freeze. Dip and
1:38
stay below minus1 and you rip. Rise
1:40
towards zero and then turn attractive
1:42
and you crunch. Three endings, one
1:45
decimal place. How space actually
1:47
stretches. You are not sitting inside a
1:50
balloon that pops. You are sitting
1:52
inside a volume that keeps adding
1:53
volume. Every mega parseek of empty
1:56
space about 3.26 million lightyears
1:59
currently gains about 70 km of extra
2:01
stretch every second. Bound things do
2:04
not feel that your coffee cup does not
2:06
expand. Earth does not expand. The solar
2:09
system does not expand. The Milky Way
2:11
does not expand. Even the local group is
2:14
glued by gravity tighter than the
2:15
stretch. Andromeda is still about 2.5
2:18
million lighty years away and still
2:20
falling toward us at a couple hundred
2:22
kilometers per second. That collision is
2:25
still on the calendar in about 4 billion
2:27
years. Expansion is a logistics problem
2:29
first. Galaxies that are not already
2:31
bound keep sliding off your map. The
2:34
ones that are bound keep their
2:35
addresses. The night sky you lose is the
2:38
night sky that was never gravitationally
2:40
yours. The freeze path. If W holds at
2:44
minus1, dark energy density stays fixed
2:47
while matter thins out. Dark energy
2:50
already owns about 68% of the energy
2:53
budget. Ordinary atoms are about 5%.
2:56
Dark matter about 27. As cubes of space
2:59
get larger, matter's share shrinks, and
3:02
dark energy share grows toward 100%.
3:05
Distant galaxies redshift until their
3:08
light stretches past any detector you
3:10
could build. The observable universe
3:12
today has a radius of about 46 billion
3:15
lighty years. In a freeze, the sphere of
3:18
stuff you can still exchange signals
3:20
with stops growing in any useful way.
3:23
You keep the local group. You lose
3:25
everything else. Stars still burn.
3:28
Planets still orbit. But the catalog of
3:30
neighbors shrinks to a handful. Then to
3:33
a merger product, then to a single
3:35
elliptical sitting in a black ocean.
3:38
That is the heat death people mean. And
3:40
it is the default if the cosmological
3:42
constant is real. Last light walk the
3:45
freeze on a longer clock. In about 100
3:48
trillion years, star formation ends. The
3:51
era of shining stars is over. White
3:54
dwarfs, neutron stars, and black holes
3:57
inherit the dark. If protons decay, even
4:00
those corpses evaporate on time scales
4:02
around 10 to the 40 years. After that,
4:06
black holes run the universe. A black
4:08
hole with the mass of the sun takes on
4:10
the order of 10^ the 67 years to
4:13
evaporate by Hawking radiation. The
4:15
biggest super massive black holes last
4:18
until about 10 to the 100 years. After
4:21
that, you have a thin fog of photons,
4:24
nutrinos, electrons, and posetrons
4:27
stretched so far that the sky is black
4:29
in every direction you point. Fred Adams
4:32
and Gregory Laughlin laid that sequence
4:35
out in reviews of modern physics in
4:37
April 1997. They were not guessing the
4:40
weather. They followed fuel, then
4:43
particles, then Hawking radiation down
4:45
to the last watt. Heat death is not an
4:48
explosion. It is a universe that can no
4:50
longer do work because every gradient
4:53
has been spent. Phantom energy cross W
4:56
below minus one and the logistics
4:58
problem becomes a physics problem.
5:01
Phantom energy gets denser as space
5:03
expands. More density means faster
5:06
expansion. Faster expansion makes more
5:09
phantom energy. On August 13th, 2003,
5:13
Robert Caldwell, Mark Kamyowski, and
5:16
Nevin Weinberg published that runaway in
5:19
Physical Review Letters. Their worked
5:21
example used w= minus1.5 and a hubble
5:25
rate of 70 km/s per mega parc time left
5:30
about 22 billion years. That is not a
5:33
measurement of our universe. It is what
5:34
happens if that value holds and never
5:37
climbs back. Later plank data put w near
5:41
minus 1.03. If the true value sits on
5:44
the phantom side of that error bar, the
5:46
rip slides out to something like 200
5:49
billion years. The direction of the
5:51
ending does not change. The date does.
5:54
You still lose every bound system. You
5:56
just lose them later. The part that
5:59
actually matters. And it gets worse
6:01
because a RIP does not wait until the
6:03
last second to start taking things
6:05
apart. In the Caldwell example, galaxy
6:08
clusters unbind about a billion years
6:11
before the end. The Milky Way is
6:13
stripped about 60 million years before.
6:15
The solar system unbinds about 3 months
6:18
before. Earth itself comes apart about
6:20
30 minutes before. Atoms last until
6:23
roughly the final 10 to the minus19
6:25
seconds. You would not see a wall of
6:27
fire coming over the horizon. You would
6:30
watch bound systems fail in order.
6:32
Largest first as the stretch outruns
6:35
every force that holds them. Gravity
6:37
loses first, then electromagnetism, then
6:40
the strong force. There is no debris
6:42
field left to inherit. Distances go to
6:45
infinity in finite time. That is the
6:47
part that actually matters. A freeze
6:50
leaves you a cold address. A rip leaves
6:52
you no address. The crunch comes back.
6:56
The third ending is the one textbooks
6:58
buried after 1998 when two supernova
7:01
teams showed the expansion was
7:03
accelerating. A crunch needs gravity to
7:06
beat the stretch. That only happens if
7:09
dark energy weakens, hits zero, and then
7:12
turns attractive. or if a negative
7:14
cosmological constant was hiding under a
7:17
temporary push. On April 4th, 2024, the
7:21
dark energy spectroscopic instrument
7:24
released its firstear barriion acoustic
7:26
oscillation results from a map of
7:29
millions of galaxies. Combined with the
7:31
cosmic microwave background and
7:33
supernova cataloges, the FIT slightly
7:36
preferred a dark energy that evolves
7:38
instead of sitting nailed at minus one.
7:41
In 2025 and again in late August 2026,
7:45
Henry Thai and collaborators ran one
7:47
such model forward. The scale factor set
7:50
to one today peaks at about 1.69 times
7:53
its present size 11 billion years from
7:56
now. The Hubble rate hits zero. Then
7:59
contraction occupies about a quarter of
8:01
the model's total lifetime. The crunch
8:04
arrives about 20 billion years from now.
8:07
That is a model consistent with a hint,
8:09
not a verdict. The hint is what put the
8:11
crunch back on the board after 20 years
8:14
of being told it was dead. What does not
8:17
happen? What does not happen tomorrow
8:19
morning? The sun does not go out. The
8:21
Hubble rate does not jump. Earth does
8:23
not unbind. You do not wake up to a
8:26
colder microwave background you can feel
8:28
with your skin. Andromeda is still
8:30
coming on a 4 billionyear clock. Proton
8:33
decay, if it happens at all, is not a
8:35
weather event. A vacuum decay bubble, if
8:38
our vacuum is even metastable, would
8:40
expand at light speed, which means you
8:42
would not get a warning and also means a
8:44
bubble that started far enough away
8:46
never reaches you. None of the three
8:49
endings is a next decade story. The desi
8:52
preference is a few sigma, not a
8:54
discovery. Plank is still consistent
8:56
with a constant. The Cornell crunch is
8:59
one best fit path through noisy data.
9:02
Your lifetime, your species lifetime,
9:04
even the remaining main sequence life of
9:06
the sun sit inside a rounding error on
9:09
all three clocks. The universe is not
9:12
ending in your inbox. Which clock you
9:15
are on, you are on one of those three
9:17
clocks right now. The freeze is the long
9:20
fade. Last stars at 100 trillion years,
9:23
last black holes at 10 to the 100. The
9:26
rip is a finite time singularity. If
9:29
phantom energy holds, the crunch is a
9:32
reversal. If dark energy dies and
9:34
gravity gets the last word, the
9:36
measurement that sorts them is not a
9:38
photograph of the end. It is w rided
9:41
from how fast space stretched when the
9:43
universe was half this age and how fast
9:46
it stretches in the maps we are still
9:48
taking. Desi is still mapping. Uklid is
9:51
still mapping. The next supernova
9:53
catalogs.
#News