Artificial intelligence can imitate any voice after 3 seconds

Artificial intelligence (AI) and other software can mimic voices that have been around for a while. But that normally requires a lot of source material for the software to learn exactly how someone pronounces each sound. But the new VALL-E from Microsoft only takes three seconds.

After such a short and random sound fragment of a voice, VALL-E can imitate the sound, pitch and pronunciation of a voice. The AI ​​was put to the test by the Cornell UniversityThat some examples shares.

Some of these examples seem natural and convincing, others seem as robotic as old attempts at artificial intelligence.

Not yet publicly available

VALL-E was formed with 60,000 hours of voice recordings. Like other AIs, VALL-E only gets better as more hardware becomes available. VALL-E is not yet publicly available.

This kind of AI is controversial because it is easy to use and especially to abuse. Anyone can thus generate a statement indistinguishable from someone else

ChatGPT in Word and Powerpoint

Microsoft, meanwhile, plans to release the AI ​​text ChatGPT to incorporate into its products. ChatGPT is a chatbot that provides natural language answers, also in Dutch.

ChatGPT will be integrated with Office Word, Powerpoint and Outlook products Information. The software could, for example, soon be able to compose emails in the tone of the user.

Microsoft also wants the chatbot to its Bing search engine to give users faster answers to their questions. With the AI ​​function, Microsoft would like to intensify the fight against Google.

Check Also

Start Spending Less on Your Everyday Purchases

Shop Smart: Start Spending Less on Your Everyday Purchases

Remember the times when our grandmas used to clip coupons from newspapers and magazines to …

Leave a Reply

Your email address will not be published. Required fields are marked *