75 years in a day of Economist Q&A since 1951 with Neumann , Einstein, Turing
40 YEARS MEDIA CHARTERING
London Celebrating AI & Quantum & 6G & Energy --- thanks to lead mapmakers Jensen & Demis & Charles3 .. EJ : : Japan+63 ... MEIM : Millennials Energy Intelligence Mapping
Old home page. -please use web version of our timeless blogs
www.economistdiary.vom Is English Language Modelling intelligent enough to sustain our human species? Great are 1990s Valley startups eg nvidia, musk's and googles exponentially linking much of whats humanly possible with machines engineered billion times mo(o)re maths brainpower than individuall human minds. But 1943 UK future shocks to.o. Geoffrey Crowther Economist Ed started debate keynes: were engineers deeper than economists in locking in futures next gens connect?. 1943 also saw dad norma cambridge studies interrupted serving last days as teen navigator allied bomber command burma. Surviving joyfully hired 1948 by Crowther to mediate engineers like Neumann Einstein Turing & Economist purpose. 3 generations apart, imost unfortunate Neumann-Einstein-Turing all left earth by early 1957: last coding notes Neumann's Computer and the Brain. Economist IQuiz disliked by EU but what to do with billion times more machine brainpower celebrated by Kennedy, & the royal families of UK & Japan. Whence not surprising greatest UK AI startups deep mind & arm influenced by royal societies & Cambridge business park ( crown property) & crick/watson open sources of dna, & cavendish lab 1920s influencing Taiwan's tech grandfather. see part 2 2025report 40 years in inteligence war between bad media and good education agents Countries with good data sovereignty projects rsvp chris.macrae@yahoo.co.uk UK Japan Taiwan India France Saudi UAE Korea S Singapore HK US 1 2

Friday, December 31, 1982

Q&A2 if your mother tongue is one of next 990 languages will LLMs and new AI put you at a disadvantage

 .


It is possible that LLM gateways to 2020s digital disadvantage people whose mother tongue is one of the next 990 languages after the 10 most powerful ones. This is because LLMs are typically trained on large datasets of text and code in the most common languages, such as English, Chinese, and Spanish. As a result, they may be better at understanding and generating text in these languages than in less common languages.

However, there are a number of things that can be done to mitigate this disadvantage. For example, researchers are developing LLMs that are specifically trained on smaller datasets of text and code in less common languages. Additionally, there are a number of tools and resources that can be used to translate text between different languages, which can help people whose mother tongue is not one of the most common languages to access and use LLM-powered applications.

Here are some specific things that can be done to mitigate the digital disadvantage of people whose mother tongue is one of the next 990 languages after the 10 most powerful ones:

  • Develop LLMs that are specifically trained on smaller datasets of text and code in less common languages. This would make LLMs more accessible to people whose mother tongue is not one of the most common languages.
  • Develop tools and resources that can be used to translate text between different languages. This would help people whose mother tongue is not one of the most common languages to access and use LLM-powered applications.
  • Provide training and support to people whose mother tongue is not one of the most common languages on how to use LLMs. This would help people to overcome the digital disadvantage that they may face.

It is important to note that the digital disadvantage is not just a problem for people whose mother tongue is not one of the most common languages. It is also a problem for people who live in rural areas, who have low incomes, or who have disabilities. We need to work to ensure that everyone has access to the digital tools and resources that they need to succeed in the 21st century.

No comments:

Post a Comment