Knowledge doubling--is there any way this isn't nonsense?

I don’t know how, even theoretically, a single atom could encode all the knowledge generated by humans by January 1, 2024. But that’s not the scenario. The scenario is that every atom in the universe is harnessed to encode more than a Googol times the knowledge generated by humans by that date.

We’re gonna need a bigger universe.

Wouldn’t that make the factoid more about communication than knowledge?

For other parts of this topic, I question the meaning of the word “knowledge.” There is a lot of stuff entered into computers (as the most common transmission medium now, I suppose) that is very far from known and is more likely made up. Sometimes I think the epistemological state of human knowledge has more decreased than otherwise.

Please - there’s more information in the DNA in my body alone than on every computer on the planet put together. Twice that if you include my gut flora.

Most of it is just copies of the same DNA, though, which at like three billion base pairs clocks in at something like a couple GB.

And what portion of all digital knowledge is unique?

How should I know? I’m not defending the claim in the OP, just pointing out that the information content in the cells of the human body would be quickly eclipsed. Besides, information isn’t the same thing as knowledge, so I’m not sure how to make the comparison anyhow.

Patent nonsense.
Somewhere we have an exponent on our exponent of growth, and maybe worse. At the rates suggested there appears to be a real danger that the Earth will collapse down into a black hole sometime in the next decade or two - simply due to the amount of information held.

The truth about any exponential growth is it can continue that way … for awhile. Then the growth has to slow to something far below exponential. Lest you get to @Francis_Vaughan’s cited end game real soon.

Lots of dumb clickbait, and lots of dumb marketing materials, are written taking the legit early exponential growth phase of something new-ish, then extrapolating that without bound. With no mention of the necessity for that growth formula to change drastically real soon.

Running the numbers, the total size in all cells in a person’s DNA in a human is only about twice the current total computer storage. And it is phenomenally redundant. The actual real data in a person’s dna in total is only about 740 MBs. It’ll fit on a CD. And that’s without using the most basic sort of compression.

The OP’s factoid can be traced to studies measuring the amount of data created and stored each year. Data is not knowledge.

The fastest annual growth was between 2010 and 2011, when the amount of data grew 150%. Which is a lot! But it’s not doubling every 12 hours. Currently the amount of data is growing by about 25% per year.

Thank you, that’s helpful–and it’s key to distinguish between “data created” and “data existing.”

If data is wealth, and if I’ve earned a million bucks a year for the past 2,000 years, I’ve got two billion bucks. If this year my earnings increase by 150%, that means this year I earn 2.5 million bucks, more than double what I earned last year. But my wealth hasn’t come anywhere close to doubling: rather, it’s increased by 0.125% instead of the expected 0.05%.

At that rate, although my income has more than doubled, my wealth will double every 800 years.

Yes, the distinction between stocks and flows is helpful to bear in mind. Also, data isn’t knowledge. Half of all data created is video, which is a rather inefficient mechanism for knowledge generation for example.

“Ah, good,” said Sedak to T’Pong, “Those primates were getting dangerously presumptive in their need to instill human values across the Alpha Quadrant with their ‘Federation of Planets’.”

“Yes,” replied T’Pong, “they were quite pernicious and oblivious to the ironic virtues of spreading ‘human rights’ on other cultures. Overwhelming them with a surfeit of spurious data which would thermodynamically collapse their planet out of physical existence was a brilliant plan, Sedak! I find your ideas intriguing and would like to join your subspacestack.”

“T’Pong, Let us dance to our success in once again protecting the galaxy from a calamitous species convinced of their predestination to rule over all, T’Pong!”

{celebratory frolicking to the beat of “Get Down Tonight” by KC & The Sunshine Band ensues}

Stranger

You rang?

Moore’s Law, strictly speaking, has been dead for a while, more from economics than technology, since it is way expensive to build new fabs to keep it going. The new trend is to put chiplets on substrates so the entire system grows while the technology of the individual chip doesn’t.

Now, for the interesting thread. If you measure knowledge by the amount of storage available, I suspect the assertion is true. I bought basically a toy camera for my five-year old grandson last Christmas. It came with a gigabyte memory card. It was not expensive. That was more memory than there was in the world when I was in college, I suspect, and was even a serious amount of memory when I worked at Intel. Now it is a toy.
I didn’t notice a good definition of knowledge here. I’ve been playing with AIs for story generation. I’ve been saving the stories in a big Libre Office file, and over just a few weeks of not a lot of work I generated 400 pages. Much of it is different versions of the same story, I doubt any of it is really knowledge, but that would be a year of work for anyone but Isaac Asimov or the reverend Fanthorpe. So in the loose sense of knowledge it might well be doubling.

How long ago did you go to college??? Up thread, I mentioned classic 9-track tapes. They could hold up to 175MB. So 6 such tapes could hold over a gigabyte. It’d be a pretty poor late 50s/early 60s data center that didn’t at least one order of magnitude more than that which would cover even low density tapes. A single major business could easily be in the terabyte range.

Asimov claimed to work alternative between three or four typewriters at a time, producing 1-3 pages a day per day, or somewhere around words in English. This is about 8k to 10k bytes. How much of this is actually original ‘knowledge’ is questionable, and certainly in terms of digital volumes of data produced today, only a tiny fraction of a percent could be viewed as unique ‘knowledge’ in any sense of the world since so much of it is repetition, but I think Asimov probably represented the technology of knowledge capture threshold circa 1960.

Stranger

Consider, for example, that the number of YouTube videos is 4.8 billion. The number of MB in one YouTube video that’s a minute long is 50,000,000. So if the average YouTube video is one minute long, there are (4.8 billion) times 50,000,000 times 8 bits ln it. (The 8 is there because a byte is 8 bits.) This there are about (48 times 5 times 8) times 1,000,000,000,000,000), which is 240 times 8, which is well over 1,000,000,000,000,000,000 bits held by YouTube. One estimate is that there have been 129,864,880 different books ever printed. A book has, say, 500 pages with 40 lines per page with a hundred characters per line with each character being specified by 10 bits. So each different book holds 20,000,000 bits of information. A set of all the books has less than 30,000,000,000,000,000 bits. So YouTube holds 30 times as many bits of information as a library with one copy each of all the books ever printed. So, yes, videos are a bigger amount of information than books.

1969 - 1973. The computer I used in high school had 4 k of disk memory.
In 1973 Ed Fredkin came to our architecture class and made the bold prediction that some day memory prices would go down to a penny a bit.
It was a different world.

So, 400 pages a year or so. Given his output, and the F&SF column and other writing, pretty reasonable. (Not counting anthologies, of course.)
I don’t know what definition of knowledge you’re using, but Asimov’s stuff is a lot closer to it than the stuff the AI pumped out. How many bytes would it take to capture all the knowledge in the world? An interesting question, but a lot less than what is online now I’d say.

Well, surely. Nobody can deny that there’s a lot of redundancy on the internet.