4 ms·

Nice writeup sama, but the history is a bit revisionist: > We are only about 30 years into the age of software, about 20 years into the age of the internet, an

Nice writeup sama, but the history is a bit revisionist: > We are only about 30 years into the age of software, about 20 years into the age of the internet, and about 2 years into the age of artificial intelligence. Software has been around--and actively making bank!--for over 40 years. The internet has been around longer in the form of academic institutions, email, and BBS than you mention. And AI has, of course, been around at least as long as software. I appreciate you have a message to send, but maybe don't ignore the path we took to get here.

criddell 10y ago
When people talk about an age, it doesn't usually start with the invention of that thing. The bronze age didn't begin with the invention of bronze. The periods are subjective, fuzzy things. I think I agree with Sam's definitions although I would probably say age of machine learning rather than age of AI.
mvp 10y ago
Those time periods are what most tech people consider as when those technologies really took off. And therefore it makes sense to say that. If he was writing a history of computer science, I agree that it would have been inaccurate to say that.
- angersock 10y ago
  "Most tech people" must mean people under the age of 25 without much knowledge of history.
- nickpsecurity 10y ago
  An example to support your view might be the switch to the trend of paying to use server resources like a utility. Centralized resource managed by people other than users that ran many parties' workloads with metered CPU, RAM, etc usage. Then, the various devices access it through a standard UI instead of native apps recoded in many OS's. Quicker setup for new users than rolling their own. Increased flexibility with the virtualization. Better physical and online security due to the centralized monitoring. On the software side, people found that safety, scalability, and maintenance were important. They needed high-level languages that did things like bounds-checks or stack protection. They needed the ability to freeze and inspect errant programs instead of crash them. They needed the software interfaces to check for right data being passed. They needed to scale on multiple CPU's. They preferred their OS's to come with source code for easier fixes or extensions. Im sure you know all about those two mainframes from IBM (1963-1970's) and Burroughs (1961) that did those things. Then came to power most large business and government institutions critical apps to this day. Anyway, I hear people made incremental improvements doing similar things with x86, Linux, and vastly lower cost-per-resource. Incremental improvements on capabilities that took off in the 60's. ;)

criddell 10y ago

When people talk about an age, it doesn't usually start with the invention of that thing. The bronze age didn't begin with the invention of bronze. The periods are subjective, fuzzy things. I think I agree with Sam's definitions although I would probably say age of machine learning rather than age of AI.

mvp 10y ago

Those time periods are what most tech people consider as when those technologies really took off. And therefore it makes sense to say that. If he was writing a history of computer science, I agree that it would have been inaccurate to say that.

angersock 10y ago
"Most tech people" must mean people under the age of 25 without much knowledge of history.
nickpsecurity 10y ago
An example to support your view might be the switch to the trend of paying to use server resources like a utility. Centralized resource managed by people other than users that ran many parties' workloads with metered CPU, RAM, etc usage. Then, the various devices access it through a standard UI instead of native apps recoded in many OS's. Quicker setup for new users than rolling their own. Increased flexibility with the virtualization. Better physical and online security due to the centralized monitoring. On the software side, people found that safety, scalability, and maintenance were important. They needed high-level languages that did things like bounds-checks or stack protection. They needed the ability to freeze and inspect errant programs instead of crash them. They needed the software interfaces to check for right data being passed. They needed to scale on multiple CPU's. They preferred their OS's to come with source code for easier fixes or extensions. Im sure you know all about those two mainframes from IBM (1963-1970's) and Burroughs (1961) that did those things. Then came to power most large business and government institutions critical apps to this day. Anyway, I hear people made incremental improvements doing similar things with x86, Linux, and vastly lower cost-per-resource. Incremental improvements on capabilities that took off in the 60's. ;)