Semiwiki on LinkedIn Semiwiki on Facebook Semiwiki on Twitter Register / Log in 
Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/threads/is-ai-the-automobile-of-this-generation.25101/
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2031070
            [XFI] => 1060170
        )

    [wordpress] => /var/www/html
)

Is AI the "automobile" of this generation?

B

Barnsley

Well-known member
Some fella on Linkedin in said AI is the automobile of this generation in that the automobile replaced the horse in the economy.

He say AI gonne replace human as an analogy.

However I ask at what stage does the cost of AI come down?

Who is going to be the Japanese car industry of AI?

Where can the disruption come from?
 
Last edited:
Barnsley said:
Some fella on Linkedin in said AI is the automobile of this generation in that the automobile replaced the horse in the economy.

He say AI gonne replace human as an analogy.

However I ask at what stage does the cost of AI come down?

Who is going to be the Japanese car industry of AI?

Where can the disruption come from?
Click to expand...

I have heard AI called the next industrial revolution and I have to agree with that.

The United States is commonly described as having gone through several major industrial/technological revolutions. Historians divide them slightly differently, but this is the standard framework:
  1. First Industrial Revolution (late 1700s–mid-1800s)
    • Water power and steam power
    • Textile mills, mechanized manufacturing, canals, railroads
    • Key figures: Eli Whitney, Samuel Slater
  2. Second Industrial Revolution (late 1800s–early 1900s)
    • Electricity, steel, oil, mass production
    • Telegraph, telephone, automobiles
    • Rise of large corporations and assembly lines
    • Key figures: Thomas Edison, Henry Ford
  3. Third Industrial Revolution / Digital Revolution (mid-1900s–early 2000s)
    • Electronics, computers, semiconductors, automation
    • Internet and telecommunications
    • Key companies: Intel, IBM
  4. Fourth Industrial Revolution (2000s–present)
    • Artificial intelligence, robotics, cloud computing, biotech, IoT
    • Fusion of physical and digital systems
    • Technologies include autonomous vehicles, generative AI, advanced semiconductors
    • Key companies: NVIDIA, OpenAI
 
Daniel Nenni said:
I have heard AI called the next industrial revolution and I have to agree with that.

The United States is commonly described as having gone through several major industrial/technological revolutions. Historians divide them slightly differently, but this is the standard framework:
  1. First Industrial Revolution (late 1700s–mid-1800s)
    • Water power and steam power
    • Textile mills, mechanized manufacturing, canals, railroads
    • Key figures: Eli Whitney, Samuel Slater
    • Electricity, steel, oil, mass production
    • Telegraph, telephone, automobiles
    • Rise of large corporations and assembly lines
    • Key figures: Thomas Edison, Henry Ford
Click to expand...

But who is gonna be the budget AI supplier?

At the minute the costs are astronomical.

I dont believe any of those other revolutions were at these levels of relative cost.
 
Barnsley said:
But who is gonna be the budget AI supplier?

At the minute the costs are astronomical.

I dont believe any of those other revolutions were at these levels of relative cost.
Click to expand...

I personally think AI is still way overhyped for it's potential vs reality.... However, the cost trend for AI is greatly improving, by several metrics:

- Energy required per token is improving >`10X per year: The hardware is improving faster than old Moore's Law scaling, something on the order of 20-50X per generation from Hopper to Blackwell, and again from Blackwell to Vera Rubin. Note - in some scenarios the improvement is 3-5X, while others it's much larger.

- Model cost per token is improving significantly: Thanks to innovations at the model level (think Deepseek-R1 type changes), and better training, a few sources are claiming 10X per year for the last 3 years and continuing

- "Capability Level" per size is improving, i.e. less resources needed to achieve a good result: There are smaller ~ 8B-27B models now that are outperforming 70B models from just a year or two ago.

I think we need to see where we land in about 5 years to determine real cost viability. The first major budget AI provider IMO is likely not to be ChatGPT since they are trying to push the Frontier, but rather one of the "fast followers", and/or Google since they have strong incentive to retain search interest.

...

Re: the revolutions - the Industrial Revolution was made possible by the introduction of low cost energy. If AI keeps improving it's "utility vs. energy", then 'cheap energy' may again enable AI. (The Industrial Revolution was technically possible as early as the Roman Empire -- but because slave labor was so cheap, there was no incentive to invest in what was needed to make the revolution happen).
 
This is an analogy: There are many donkeys in the vicinity of Big Bend National Park. Because using heavy equipment in such rugged terrain is a one-way trip. You can get equipment in but can't get it out again.

This is complicating building a wall there. It complicates building anything. It has to be done with pack animals.

My analogy is this: The modern world can indeed be very modern but it's unevenly distributed. AI is probably not going to play a big role in about a fifth of Texas, the Big Bend West Texas region, because they need to focus more on feeding the pack animals. Humans and donkeys are safe there.
 
You must log in or register to reply here.
Back
Top