Warning: Declaration of wp_option::render($field_html, $colspan = false) should be compatible with base_wp_option::render() in /home/marcosmucheroni/www/blog/wp-content/themes/chocotheme/lib/theme-options/option-fields.php on line 123

Warning: Declaration of wp_option_text::render() should be compatible with wp_option::render($field_html, $colspan = false) in /home/marcosmucheroni/www/blog/wp-content/themes/chocotheme/lib/theme-options/option-fields.php on line 170

Warning: Declaration of wp_option_textarea::render() should be compatible with wp_option::render($field_html, $colspan = false) in /home/marcosmucheroni/www/blog/wp-content/themes/chocotheme/lib/theme-options/option-fields.php on line 194

Warning: Declaration of wp_option_choose_category::render() should be compatible with wp_option::render($field_html, $colspan = false) in /home/marcosmucheroni/www/blog/wp-content/themes/chocotheme/lib/theme-options/option-fields.php on line 226

Warning: Declaration of wp_option_choose_page::render() should be compatible with wp_option::render($field_html, $colspan = false) in /home/marcosmucheroni/www/blog/wp-content/themes/chocotheme/lib/theme-options/option-fields.php on line 240

Warning: Declaration of wp_option_select::render() should be compatible with wp_option::render($field_html, $colspan = false) in /home/marcosmucheroni/www/blog/wp-content/themes/chocotheme/lib/theme-options/option-fields.php on line 257

Warning: Declaration of wp_option_file::render() should be compatible with wp_option::render($field_html, $colspan = false) in /home/marcosmucheroni/www/blog/wp-content/themes/chocotheme/lib/theme-options/option-fields.php on line 301

Warning: Declaration of wp_option_separator::render() should be compatible with wp_option::render($field_html, $colspan = false) in /home/marcosmucheroni/www/blog/wp-content/themes/chocotheme/lib/theme-options/option-fields.php on line 408

Warning: Declaration of wp_option_choose_color_scheme::render() should be compatible with wp_option::render($field_html, $colspan = false) in /home/marcosmucheroni/www/blog/wp-content/themes/chocotheme/lib/theme-options/choose-color-scheme.php on line 39

Warning: Use of undefined constant wp_cumulus_widget - assumed 'wp_cumulus_widget' (this will throw an Error in a future version of PHP) in /home/marcosmucheroni/www/blog/wp-content/plugins/wp-cumulus/wp-cumulus.php on line 375
fevereiro « 2019 « Blog Marcos L. Mucheroni Filosofia, Noosfera e cibercultura
RSS
 

Arquivo para fevereiro, 2019

[:pt]Até onde pode ir a AI[:en]Even waves can go to AI [:]

20 fev

[:pt]O projeto OpenAI que embora se diga “sem fins lucrativos” e seja realmente aberto, basta entrar no blog do projeto para verificar o andamento e as possibilidades, na verdade assuntou até mesmo os especialistas ao publicar um sistema que escreve notícias/textos, teoricamente de ficções, mas que podem ser classificados como fake, ou como estão sendo chamados: faketextos.
Também o código está aberto e disponível no site GitHub para desenvolvedores.
Os sistemas de processamento de linguagem natural podem fazer tarefas como responder questões, conversão de máquinas, compreensão de leitura e sumarização de textos, o que já são tipicamente abordadas com aprendizado supervisionado em conjuntos de dados específicos de tarefas, porém a pesquisa de textos no GPT2 é mais ampla em quantidade.
O GPT2, sucessor do GPT que era apenas um produtor de textos a partir de textos básicos, agora pode ler até 40 GB de textos existentes na Web, e o que tem produzido assusta um pouco, pela clareza, profundidade e o pior de tudo, pura ficção ou mais claramente: fakes.
Entre suas características sintáticas, é superior a outros do género, escrevendo passagens que fazem sentido com o texto anterior e mantém o estilo, sem se perder em frases longas.
O problema é que poderá gerar fakenews que agora podem ser mais longas, tornando-se faketextos, uma reportagem do The Guardian mostra a novidade e os problemas:

 [:en]The OpenAI project, although it is said to be “non-profit” and is really open, just enter the blog of the project to check the progress and the possibilities, in fact even pointed the experts when publishing a system that writes news / texts, theoretically fictions, but that can be classified as fake, or as they are being called: faketexts.  Also the code is open and available on the GitHub site for developers. 
Natural language processing systems can perform tasks such as answering questions, machine conversion, reading comprehension, and text summarization, which are already typically addressed with supervised learning in task-specific data sets, but text search in GPT2 is wider in quantity.
GPT2, the successor to the GPT that was just a producer of texts from basic texts, can now read up to 40 GB of existing text on the Web, and what it has produced frightens a little, for clarity, depth and worst of all, pure fiction or more clearly: fakes.
Among its syntactic characteristics, it is superior to others of the sort, writing passages that make sense with the previous text and maintains the style, without getting lost in long sentences. The problem is that it can generate fakenews that can now be longer, becoming faketexts, a report from The Guardian shows the novelty and the problems:

[:]