Beep boop - this is a robot. A new show has been posted to TWiT…
What are your thoughts about today’s show? We’d love to hear from you!
I rarely full screen anything. I always like to have my windows tiled, and have at least some pixels of the desktop showing at all times. I usually use fancyzones even on laptops (like at work).
I usually have things come out from the top corner left corner, and bottom right corner.
2 Likes
The same here. When my laptop is docked, it is connected to a 4K UHD monitor and that is too big for maximising windows (I run at 100% scaling). In the move, I also don’t tend to maximise windows.
1 Like
Regarding AI being open, that is no longer the case. Until the mid to late 2010s, a majority of AI research was being done by universities, often with sponsorship, and the research papers were open, but there was a move to restrict that research, or rather a move to restrict the publishing of papers based on that research.
By 2020, AI research papers had dropped by 20%, a lot of key researchers were moving from open research in institutions to closed research in companies. The elite universities were severely hit in the USA and China saw a boom in AI research papers, compared to the US during that time, but that seems to be going the same way, as far as I can see.
Researchers working in deep-learning are the most highly impacted, with most influencial researchers now working directly for industry. Leo covered it several times over the years on TWiT and TWiG, so I was surprised he was the one saying that the research is open.
Also, those companies have gagged researchers or fired them, if they try and release papers that point out inconvenient facts - E.g. Timnit Gebru, who was a co-lead in the Google Brain. She wrote a paper for an AI conference about the limitations and risks of LLMs and concerns about the fairness of models trained on “noisy Internet data”, their substantial environmental impacts and their limited ability to understand language as compared to generate plausible-reading text.
Regarding the use of AI for dubbing film, I’ve heard some NotebookLM “podcasts” and I’ve seen AIs trying to translate and what a poor job they do, compared to professionals. In Germany the dubbing industry is huge and generally well respected. People recognise the voice artists, many good artists will be the voices for several actors and you will recognise their voices in different films, for example Dietmar Wunder, the voice of Daniel Craig and Don Chiedal, among others or Tobias Kluckert (Joaquin Phoenix, Gerard Butler, Guillaume Depardieu, Tyrese Gibson, Nathan Fillion and Karl Urban, among others). They are real actors who put real feelings into the dubbing, in some cases the dubbed roles are better than the originals, IMHO.
Also, the dubbing is not just a “translation” of the original text, it is substituting some words or phrases, so that the lipsync works most of the time, if you aren’t a lip reader.
I very much doubt that AI can provide the same level of characterisation and feeling, let alone a clean translation. I am willing to be proven wrong.
1 Like