AL introduces AI voice system for lawmakers’ speech translations

The Legislative Assembly (AL) has introduced a new method to broadcast the Portuguese translations of lawmakers’ speeches and inquiries, particularly during the period before the agenda.

In yesterday’s plenary session, the traditional speech-reading by several AL Portuguese-speaking staff members was replaced with a synthesized voice that covered all the lawmakers’ pre-agenda speeches, except for José Pereira Coutinho, who traditionally delivers such speeches and inquiries in Portuguese.

Contrary to what happened in the other speeches, in this case, the Chinese translation of the speech was read aloud by a Cantonese-speaking AL staff member without the use of Artificial Intelligence tools or computer-assisted reading.

Changes in the AL’s approach to parts of its work had already been noticed during the most recent visit of the Chief Executive (CE), Sam Hou Fai, to a Q&A session with lawmakers on the policy address for the second semester of this year, on June 16.

At the time, the Times reporter also noticed that the CE’s speech in Chinese was accompanied by automatic Chinese subtitles, something that had not happened before.

Although the AL has not yet issued any clarification or official statement regarding these novelties, the president of the AL, André Cheong, after being elected to this post, said that one of his major tasks would be to focus on modernizing the AL’s operations and procedures.

Since mid-October last year, when Cheong was elected to the post, the AL has passed several resolutions aimed at amending procedures, including lawmakers’ speech durations and other aspects, which Cheong noted had been amended to increase efficiency.

Previously, with the inception of the Sam Hou Fai government, the sectoral policy address debates were amended from a single session starting at 3 p.m. and ending at midnight (12 a.m.) to a double session starting at 9:30 a.m. and concluding at 8 p.m., with a lunch break.

For the time being, the synthesized voice is used only for lawmakers’ speeches, not for government presentations, which continue to be translated and read by people.