The evolution of predictive text is an example of how machine learning has developed under our own eyes, without us really noticing how powerful, and accurate, it has become.

When it first started, it could pretty much complete a word, but it had to almost see a majority of the letters typed to offer up what it thought you were typing. It was often wrong, but got better as the technology improved.

It then developed into predicting what the next word would be, based on how a simple sentence is structured. Then a more complex sentence based on the previous words used in the sentence.

Then it became more personalised to the user, by building a dictionary of commonly typed words to suggest.

Then there was the feature that learnt from the user's writing style, by analysing texts and emails.

I've had the same Android phone for the last six years. It's scanned multiple emails and texts, and has built up a dictionary so personal, it knows pretty much every word I use. All my nuances, slang, swear words, brand names, beer styles, rugby teams, places. When I start a text to certain people, it pretty much knows the tone of the text by multiple words in advance, and can almost complete the sentences for me.

I haven't programmed it in the traditional sense of programming. It's just learnt from repetition over a period of time.

This is an interesting read regarding the future of predictive text.

https://www.newyorker.com/magazine/2019 ... new-yorker

The evolution of predictive text is an example of how machine learning has developed under our own eyes, without us really noticing how powerful, and accurate, it has become.

When it first started, it could pretty much complete a word, but it had to almost see a majority of the letters typed to offer up what it thought you were typing. It was often wrong, but got better as the technology improved.

It then developed into predicting what the next word would be, based on how a simple sentence is structured. Then a more complex sentence based on the previous words used in the sentence.

Then it became more personalised to the user, by building a dictionary of commonly typed words to suggest.

Then there was the feature that learnt from the user's writing style, by analysing texts and emails.

I've had the same Android phone for the last six years. It's scanned multiple emails and texts, and has built up a dictionary so personal, it knows pretty much every word I use. All my nuances, slang, swear words, brand names, beer styles, rugby teams, places. When I start a text to certain people, it pretty much knows the tone of the text by multiple words in advance, and can almost complete the sentences for me.

I haven't programmed it in the traditional sense of programming. It's just learnt from repetition over a period of time.

This is an interesting read regarding the future of predictive text.

https://www.newyorker.com/magazine/2019 ... new-yorker

I read those books and it struck me that if the development of AI was in the hands of benevolent actors, that is more than likely what they would aim for; it was refreshing to explore a utopian future, rather than the dystopia that is often the default of sci-fi writers. Cory Doctorow has also written some interesting stuff around post-scarcity anarchy, and the future of society and how it interacts with technology.

To the question - I would happily live as a Culture citizen:

"The Culture's economy is maintained automatically by its non-sentient machines, with high-level work entrusted to the Minds' subroutines, which allows its humanoid and drone citizens to indulge their passions, romances, hobbies, or other activities, without servitude. Biologically, the Culture's citizens have been genetically enhanced to live for centuries, and have modified mental control over their physiology, including the ability to introduce a variety of psychoactive drugs into their systems, change biological sex, or switch off pain at will. Culture technology is able to transform individuals into vastly different body forms, although the Culture standard form remains fairly close to human."

Works for me.

I read those books and it struck me that if the development of AI was in the hands of benevolent actors, that is more than likely what they would aim for; it was refreshing to explore a utopian future, rather than the dystopia that is often the default of sci-fi writers. Cory Doctorow has also written some interesting stuff around post-scarcity anarchy, and the future of society and how it interacts with technology.

To the question - I would happily live as a Culture citizen:

"The Culture's economy is maintained automatically by its non-sentient machines, with high-level work entrusted to the Minds' subroutines, which allows its humanoid and drone citizens to indulge their passions, romances, hobbies, or other activities, without servitude. Biologically, the Culture's citizens have been genetically enhanced to live for centuries, and have modified mental control over their physiology, including the ability to introduce a variety of psychoactive drugs into their systems, change biological sex, or switch off pain at will. Culture technology is able to transform individuals into vastly different body forms, although the Culture standard form remains fairly close to human."

Works for me.