Long before most of you were born there was a special number you could call to get the time here in the USA. It would give you a localized time every 10 seconds. It was considered an authoritative value, but I knew that normal analog clocks drifted a fair amount, and I was fascinated by the Time voice.
There was also a special number to dial that was called Information. When I was about 12, I suddenly thought that’s amazing, you can just call and get information. What I didn’t know at the time was that it only applied to phone numbers.
One day, I called Information and asked them how accurate the time number was and what mechanism was used to deliver it. Incredibly, they connected my preadolescent self to a public relations person at AT&T. That person somehow took me seriously and hooked me up with another party. And that person told me about the cesium atomic clock used to keep track of the time.
Later, I would call Information to find out the acreage occupied by the Los Angeles Zoo, and it was then that they brusquely explained to me that the number was not for general knowledge.
Also, where I live, radio transmitted the precise time in the most low tech way possible. A person said: "It is XX:XX time with the beep" and this was followed by four beeps, three short, one long, with the end of the last beep signifying the exact moment of the announced time.
For the kids out there, calling a number for recorded information was common in the 90s I guess? Movie times was a big one. Where I was locally, multiple surf shops did 3-4x daily surf reports. The later was a bit funny because if one of the usual update times was missed you knew the waves were good and the guy was surfing :)
In Northern California the number corresponded to POP-CORN, so if you wanted to know the time you called popcorn, and got a lovely lady's voice: "At the tone, the time will be...". The POP prefix was a shortcut to the phone company.
That's brilliant. Sometimes talking to real humans is great.
It reminds me of two things. When I was a child I was given a postage stamp and wrote a letter to "Wallace and Gromit Ltd" and posted it. Remarkably, I got a reply from Aardman Animations with a card signed by Nick Park himself. Someone at Royal Mail obviously understood and looked up the real address of the studio.
Later as a teen I noticed the clock on our oven displayed "24:00" at midnight instead of the more conventional "0:00"[0]. I wrote to the manufacturer (Neff) on a support channel and it actually got forwarded through several departments before I received a reply from an engineer quoting ISO8601 and how it doesn't actually specify that midnight has to be 0:00 (although, sadly, I never received a positive reason for why someone made the oven clock like this).
For those of us who remember such interactions it's even more frustrating when you find there's a dumb robot at the end of the line. Even if it's LLM powered. I actually asked ChatGPT what it would do with the Wallace and Gromit letter and it couldn't find the same solution. We lost a lot with Google-style faceless companies.
As for the time, in the UK we had (have?) "The Talking Clock" which you could get to by dialling 123. This was well enough known in 2001 to be a key part of this scene from The Office: https://www.youtube.com/watch?v=SdBCFV6qa3A&t=146
> Today’s best atomic clocks are billions of times more accurate and precise than any other type of clock. If they had been running since the Big Bang, 13.8 billion years ago, they would have lost or gained less than a second.
Absolutely incredible, but then confused by the next assertion that is made without context:
> New kinds of clocks may soon open up capabilities that previous generations could only dream about.
What is possible by pushing past that incredibly narrow margin of 1s in 13b years? What dreamy capabilities do we get that we don’t enjoy today?
Edit: I’d assumed they were talking about new approaches to making these clocks more accurate, but I wonder if they’re referring to power and form factors? I.E. an atomic clock the size of a grain of rice: https://www.nist.gov/noac/success-story-chip-scale-atomic-cl...
> What is possible by pushing past that incredibly narrow margin of 1s in 13b years? What dreamy capabilities do we get that we don’t enjoy today?
You can measure speed (via special relativity) and altitude (via gravity/general relativity) with atomic clocks [0], so with a more accurate clock, you should be able to calculate your speed/altitude more accurately.
GPS is already pretty good at measuring speed/altitude, but its not always very accurate, and it only works if you can receive the satellite's signal. In theory, a more accurate atomic clock would solve both of these problems (as long as you had an equally-accurate clock at a fixed location to compare to).
There are still big leaps, like their optical lattice clock that's sensitive to the gravitational difference caused by tidal effects on the earth's crust.
It can detect gravitational differences of a few cm!
Pretty sure you can rig something far better than radar if you have the ability to detect extremely tiny gravitational waves via an array of hyper-accurate clocks.
> detect extremely tiny gravitational waves via an array of hyper-accurate clocks.
I wonder how tiny is feasible. Would it get to a level of accuracy where you could detect neutrinos? We know they have some (v tiny) mass, so it should in theory be detectable that way?
In radio astronomy, more precise clocks can directly improve the quality/resolution of observations.
> Data received at each antenna in the array include arrival times from a local atomic clock, such as a hydrogen maser. At a later time, the data are correlated with data from other antennas that recorded the same radio signal, to produce the resulting image. The resolution achievable using interferometry is proportional to the observing frequency.
What, Exactly, Is a Clock?
To understand atomic time, we must first understand something about clocks.
A clock is, at its core, a marriage of two components: a mechanism that oscillates, or ticks with a steady beat, and a device that counts those beats and displays the time
Yeah but more deeply time is a property of the universe that the clock including atomic clocks are tracking. Kant said it is 'schema' i.e. without understanding time we wouldn't even understand causality so it is the purest possible a-priori concept that doesn't rely on sensory input to learn about
This is why I like using my Alphadec system that splits the year into 26 chunks from A-Z <https://github.com/firasd/alphadec> as my working dir; eg. right now the current Alphadec is '2026_P5M7_200286' so my working dir is Documents/work/2026_P. That way there is a folder system automatically imposed on various .js or .py stuff without having to pre-create a taxonomy since 'time is schema'--I can estimate "that ffmpeg project was in Spring so it must be around 2026_F.."
Not sure I get what is better about this than simply using the month, like Documents/work/202606 (which also sorts nicely and does not need a conversion in my head to know what time of the year that must be)
Yup as far as the folder taxonomy thing ISO YYYYMM works fine too
Alphadec is just a timezone free UTC encoding that works down the millisecond so I also use it for snapshots eg
myscript.py.backup.2026_P5G6_328054_Aug06.txt
So then when I'm doing multiple revisions I don't have to worry about which revision of 20260806's day it was etc. Unix time in ms would work too but Alphadec is a little bit nicer to me than saying myscript.py.backup.1786086462604.txt
Also, has anyone else ever used their Randomness Beacon? I once used it to settle an argument in a book club over which book to read next. It's basically a very fancy and high tech coin flip.
Fun fact, two (very expensive) wristwatches have actual atomic timekeeping in them, instead of just receiving a radio time signal: the Bathys Hawaii Cesium 133, and a few different shaped from Hoptroff London.
All released 2013-2015. Not sure what happened then for there to be suddenly two.
Symmetricom (now owned by Microchip) released the first Chip Scale Atomic Clocks in 2011. It probably took some time for them to become available to watch designers and for them to be integrated into watch designs.
Yeah, likewise the Breitling Emergency is big because it's an actual working 406 MHz Personal Locator Beacon. In principle that's all you need to have a rescue team come get you if you, I dunno, crashed your plane on a desert island or you broke your leg while hiking in Siberia, or whatever. Of course if you're the sort of person who wears a wristwatch which costs as much as a car you probably aren't doing anything like that, but you could be and maybe you're an elite solo who "travels light" and so you couldn't pack a conventional much cheaper PLB.
I can't imagine Breitling sells very many of these, they're huge heavy watches, if you're The Rock they probably look fine on your arm but the average tough guy can't justify owning a watch this expensive and if you're the 5'4" female CEO of a famous company you can afford the Emergency but it looks ridiculous and feels like you're carring a boulder on your wrist.
Like Concorde, sometimes we prove something was possible but it turns out it was unavoidably too expensive and annoying for it to make much sense in the market.
you would be surprised how large and passionate the watch community is. In the world of watches that breitling is big, but not really that big. It is quite a popular watch, and not that expensive : https://www.chrono24.com/breitling/emergency--mod21.htm?dose...
Sadly, no major miniaturization breakthroughs since. And those CSACs still eat up a lot of power (as do good OCXOs), making them rough for battery powered wearables.
Long before most of you were born there was a special number you could call to get the time here in the USA. It would give you a localized time every 10 seconds. It was considered an authoritative value, but I knew that normal analog clocks drifted a fair amount, and I was fascinated by the Time voice.
There was also a special number to dial that was called Information. When I was about 12, I suddenly thought that’s amazing, you can just call and get information. What I didn’t know at the time was that it only applied to phone numbers.
One day, I called Information and asked them how accurate the time number was and what mechanism was used to deliver it. Incredibly, they connected my preadolescent self to a public relations person at AT&T. That person somehow took me seriously and hooked me up with another party. And that person told me about the cesium atomic clock used to keep track of the time.
Later, I would call Information to find out the acreage occupied by the Los Angeles Zoo, and it was then that they brusquely explained to me that the number was not for general knowledge.
Also, where I live, radio transmitted the precise time in the most low tech way possible. A person said: "It is XX:XX time with the beep" and this was followed by four beeps, three short, one long, with the end of the last beep signifying the exact moment of the announced time.
For the kids out there, calling a number for recorded information was common in the 90s I guess? Movie times was a big one. Where I was locally, multiple surf shops did 3-4x daily surf reports. The later was a bit funny because if one of the usual update times was missed you knew the waves were good and the guy was surfing :)
777-FILM "Hello, and welcome to Moviefone!"
Here in Australia we had a number that gave you the time called 'George the talking clock'.
https://1194online.com
In Northern California the number corresponded to POP-CORN, so if you wanted to know the time you called popcorn, and got a lovely lady's voice: "At the tone, the time will be...". The POP prefix was a shortcut to the phone company.
https://en.wikipedia.org/wiki/Point_of_presence
https://en.wikipedia.org/wiki/Telephone_exchange_names#Stand...
411 Directory Assistance in 1978 https://youtu.be/UpsGpZ7wi9k?si=gSKP2K7WGPn-cbBj&t=60That's brilliant. Sometimes talking to real humans is great.
It reminds me of two things. When I was a child I was given a postage stamp and wrote a letter to "Wallace and Gromit Ltd" and posted it. Remarkably, I got a reply from Aardman Animations with a card signed by Nick Park himself. Someone at Royal Mail obviously understood and looked up the real address of the studio.
Later as a teen I noticed the clock on our oven displayed "24:00" at midnight instead of the more conventional "0:00"[0]. I wrote to the manufacturer (Neff) on a support channel and it actually got forwarded through several departments before I received a reply from an engineer quoting ISO8601 and how it doesn't actually specify that midnight has to be 0:00 (although, sadly, I never received a positive reason for why someone made the oven clock like this).
For those of us who remember such interactions it's even more frustrating when you find there's a dumb robot at the end of the line. Even if it's LLM powered. I actually asked ChatGPT what it would do with the Wallace and Gromit letter and it couldn't find the same solution. We lost a lot with Google-style faceless companies.
As for the time, in the UK we had (have?) "The Talking Clock" which you could get to by dialling 123. This was well enough known in 2001 to be a key part of this scene from The Office: https://www.youtube.com/watch?v=SdBCFV6qa3A&t=146
[0] I just looked this up again and, funnily enough, John Ward showed a similar oven on his YouTube channel: https://www.youtube.com/watch?v=4_oLT7lWQ44
> Today’s best atomic clocks are billions of times more accurate and precise than any other type of clock. If they had been running since the Big Bang, 13.8 billion years ago, they would have lost or gained less than a second.
Absolutely incredible, but then confused by the next assertion that is made without context:
> New kinds of clocks may soon open up capabilities that previous generations could only dream about.
What is possible by pushing past that incredibly narrow margin of 1s in 13b years? What dreamy capabilities do we get that we don’t enjoy today?
Edit: I’d assumed they were talking about new approaches to making these clocks more accurate, but I wonder if they’re referring to power and form factors? I.E. an atomic clock the size of a grain of rice: https://www.nist.gov/noac/success-story-chip-scale-atomic-cl...
> What is possible by pushing past that incredibly narrow margin of 1s in 13b years? What dreamy capabilities do we get that we don’t enjoy today?
You can measure speed (via special relativity) and altitude (via gravity/general relativity) with atomic clocks [0], so with a more accurate clock, you should be able to calculate your speed/altitude more accurately.
GPS is already pretty good at measuring speed/altitude, but its not always very accurate, and it only works if you can receive the satellite's signal. In theory, a more accurate atomic clock would solve both of these problems (as long as you had an equally-accurate clock at a fixed location to compare to).
[0]: https://news.ycombinator.com/item?id=16473776
There are still big leaps, like their optical lattice clock that's sensitive to the gravitational difference caused by tidal effects on the earth's crust.
It can detect gravitational differences of a few cm!
Pretty sure you can rig something far better than radar if you have the ability to detect extremely tiny gravitational waves via an array of hyper-accurate clocks.
> detect extremely tiny gravitational waves via an array of hyper-accurate clocks.
I wonder how tiny is feasible. Would it get to a level of accuracy where you could detect neutrinos? We know they have some (v tiny) mass, so it should in theory be detectable that way?
In radio astronomy, more precise clocks can directly improve the quality/resolution of observations.
> Data received at each antenna in the array include arrival times from a local atomic clock, such as a hydrogen maser. At a later time, the data are correlated with data from other antennas that recorded the same radio signal, to produce the resulting image. The resolution achievable using interferometry is proportional to the observing frequency.
https://en.wikipedia.org/wiki/Very-long-baseline_interferome...
> What dreamy capabilities do we get that we don’t enjoy today?
The next gen will have an alarm with snooze button.
What are these atomic clocks measured against?
I think that they typically use lots of other atomic clocks [0].
[0]: https://en.wikipedia.org/wiki/International_Atomic_Time
Not many people know NIST offers a cookbook for perfectly calibrated food
https://www.nist.gov/pml/owm/metric-si/metric-kitchen/metric...
What, Exactly, Is a Clock? To understand atomic time, we must first understand something about clocks.
A clock is, at its core, a marriage of two components: a mechanism that oscillates, or ticks with a steady beat, and a device that counts those beats and displays the time
Yeah but more deeply time is a property of the universe that the clock including atomic clocks are tracking. Kant said it is 'schema' i.e. without understanding time we wouldn't even understand causality so it is the purest possible a-priori concept that doesn't rely on sensory input to learn about
This is why I like using my Alphadec system that splits the year into 26 chunks from A-Z <https://github.com/firasd/alphadec> as my working dir; eg. right now the current Alphadec is '2026_P5M7_200286' so my working dir is Documents/work/2026_P. That way there is a folder system automatically imposed on various .js or .py stuff without having to pre-create a taxonomy since 'time is schema'--I can estimate "that ffmpeg project was in Spring so it must be around 2026_F.."
Not sure I get what is better about this than simply using the month, like Documents/work/202606 (which also sorts nicely and does not need a conversion in my head to know what time of the year that must be)
Yup as far as the folder taxonomy thing ISO YYYYMM works fine too
Alphadec is just a timezone free UTC encoding that works down the millisecond so I also use it for snapshots eg
So then when I'm doing multiple revisions I don't have to worry about which revision of 20260806's day it was etc. Unix time in ms would work too but Alphadec is a little bit nicer to me than saying myscript.py.backup.1786086462604.txtNIST is a great example of a true Public Good.
Also, has anyone else ever used their Randomness Beacon? I once used it to settle an argument in a book club over which book to read next. It's basically a very fancy and high tech coin flip.
https://csrc.nist.gov/projects/interoperable-randomness-beac...
Fun fact, two (very expensive) wristwatches have actual atomic timekeeping in them, instead of just receiving a radio time signal: the Bathys Hawaii Cesium 133, and a few different shaped from Hoptroff London.
All released 2013-2015. Not sure what happened then for there to be suddenly two.
Symmetricom (now owned by Microchip) released the first Chip Scale Atomic Clocks in 2011. It probably took some time for them to become available to watch designers and for them to be integrated into watch designs.
https://www.nist.gov/noac/success-story-chip-scale-atomic-cl...
I'm guessing the watches sold poorly and the chip packages are still large compared to a watch movement, so no new designs.
Yeah, likewise the Breitling Emergency is big because it's an actual working 406 MHz Personal Locator Beacon. In principle that's all you need to have a rescue team come get you if you, I dunno, crashed your plane on a desert island or you broke your leg while hiking in Siberia, or whatever. Of course if you're the sort of person who wears a wristwatch which costs as much as a car you probably aren't doing anything like that, but you could be and maybe you're an elite solo who "travels light" and so you couldn't pack a conventional much cheaper PLB.
I can't imagine Breitling sells very many of these, they're huge heavy watches, if you're The Rock they probably look fine on your arm but the average tough guy can't justify owning a watch this expensive and if you're the 5'4" female CEO of a famous company you can afford the Emergency but it looks ridiculous and feels like you're carring a boulder on your wrist.
Like Concorde, sometimes we prove something was possible but it turns out it was unavoidably too expensive and annoying for it to make much sense in the market.
[Edit: Fix spelling of Concorde]
you would be surprised how large and passionate the watch community is. In the world of watches that breitling is big, but not really that big. It is quite a popular watch, and not that expensive : https://www.chrono24.com/breitling/emergency--mod21.htm?dose...
Sadly, no major miniaturization breakthroughs since. And those CSACs still eat up a lot of power (as do good OCXOs), making them rough for battery powered wearables.
MEMS! You should get in touch with the folks at SiTime.