Something new to ruin his day
Not exactly shocked that a singing synth would get to this point with machine learning
Also, Google Bard taught itself Bengali without being asked to according to Googleâs AI lead
https://twitter.com/Variety/status/1648326996580073473
Dialogue Boost analyzes the original audio in a movie or series and identifies points where dialogue may be hard to hear. Then speech patterns are isolated and audio is enhanced to make the dialogue clearer. According to Amazon, the AI-based approach delivers a targeted enhancement to portions of spoken dialogue rather than a general amplification at the center channel in a home theater system.
⌠maybe they need a robot spider instead
There are increasing numbers of reports about languages learned by AI. What used to make me feel secure about services that claim ownership of audio created by consumers was the sheer volume of data. We generate so much content every day. This jumped out at me FTA:
Whisper, launched in September by the company behind the ChatGPT chatbot, was trained on 680,000 hours of audio from the web, including 1,381 hours of the MÄori language.
Indigenous tech and culture experts say that while such technologies can help preserve and revive their languages, harvesting their data without consent risks abuse, distorting of Indigenous culture, and depriving minorities of their rights.
Now, Iâm starting to think I need to expand on the method used by the Tamarians in the ST:TNG episode âDarmok.â I canât imagine the impact AI is having on codes used by intelligence agencies and military organizations.
Hm. I wonder if my site is on the list? I donât think I ever blocked the CommonCrawl bot.
In popular Twitter threads, two prominent AI researchers questioned these claims. Margaret Mitchell, a researcher and ethicist at AI startup Hugging Face who formerly co-led Googleâs AI ethics team, pointed out that PaLM, is, in fact, trained on Bengali according to a paper published by Googleâs own researchers. The paper says that Bengali made up 0.026% of PaLMâs training data.
âBy prompting a model trained on Bengali with Bengali, it will quite easily slide into what it knows of Bengali: This is how prompting works,â Mitchell tweeted. It is not possible, she added, for AI âto speak well-formed languages that youâve never had access to.â
A CBS spokesperson did not respond on the record to BuzzFeed Newsâ requests for comment.
Google first showed off PaLM at its annual conference for developers last year. Pichai himself demonstrated the softwareâs ability to understand and respond to questions in Bengali onstage.
Ew. I will not let it drive my car, it is scratched.
It does well enough through a book opening, until it tries to plow a bishop through its own pawn. Obviously it has no representation of the board or list of pieces to sanity check against. After that, the comments get more erratic, and it completely loses track by the 19th move.
It can fake well for a while, but it canât play chess.
eta: I was impressed by how well it faked it at first. It took its concentrated Internet and squeezed out what playing Chess looks like. It followed the script of moves well for a while, but since itâs not really playing Chess, once it skipped, it was lost. The lights are on, but nobodyâs home.
Hilarious
Google has assured the Financial Times that it will put firm guardrails in place to prevent errors and misinformation
Yeah, right. Theyâve never really had it for real ads, why is this any different?