Google's Latest AI Breakthrough: Generating Music from Text Prompts

The last twelve months have seen a constant stream of new types of generative AI tools coming to the internet. Some like DALL-E 2 and ChatGPT have genuinely caught the internet’s imagination a blown people’s minds. Others like Meta’s Make-A-Video tool and OpenAI’s 3D model-building tool seem more like works in progress that have failed to impress or make a splash. The latest generative AI tool coming from google seems to be something that combines both aspects. It is still rough around the edges, but there is definitely enough there to impress people already. Let's check it out:
The new tool from Google is called MusicLM. It works much like all the other generative tools we’ve seen so far in that it responds to text prompts. However, instead of giving us text, images, or videos, MusicLM will create music from text prompts alone. Not just sounds but music.
In the research paper explaining how they developed the tool, Google says:
“a model for generating high-fidelity music from text descriptions such as “a calming violin melody backed by a distorted guitar riff” MusicLM casts the process of conditional music generation as a hierarchical sequenceto-sequence modeling task, and it generates music at 24 kHz that remains consistent over several minutes.”
The results are rather startling too and quite impressive with lengthy beats being composed from simple text prompts “melodic techno” and “relaxing jazz”. Things do get really weird though when the music has what seems to be AI-generated vocals, which can come across as rather chilling. You can check them out some of the music created from various prompts below and check out all the available sound clips here:
“The main soundtrack of an arcade game. It is fast-paced and upbeat, with a catchy electric guitar riff. The music is repetitive and easy to remember, but with unexpected sounds, like cymbal crashes or drum rolls.”
“A rising synth is playing an arpeggio with a lot of reverb. It is backed by pads, sub bass line and soft drums. This song is full of synth sounds creating a soothing and adventurous atmosphere. It may be playing at a festival during two songs for a buildup.”
“This is an r&b/hip-hop music piece. There is a male vocal rapping and a female vocal singing in a rap-like manner. The beat is comprised of a piano playing the chords of the tune with an electronic drum backing. The atmosphere of the piece is playful and energetic. This piece could be used in the soundtrack of a high school drama movie/TV show. It could also be played at birthday parties or beach parties.”
Clearly the models still have a long way to go before they will be creating high-quality music that will attract many users. However, Google claims that Music LM is head and shoulders above other generative music models, and it has to be said that the outputs showcased are pretty compelling when you consider that they came from just a few lines of text. Just don't tell Nick Cave.
At this point, Google has said that it doesn’t have any plans to release these models to the public, which means you’ll have to stick to ChatGPT if you want to scratch your generative AI itch.


“Do you use Google Photos?”
I do; I find it impossible not to use Google Photos on the Android phone; nevertheless, the “memory” feature is sort of neat. I’ve seen photos from a couple of years ago that that offer glimpses into the long-ago, forgotten past. It’s a lot like reviewing journal writing. “What was I doing and such and such a date?”
And, I think, when the “memories” are sorted and positioned, one can create a mini-collage with up to eight photos.
It’s so much easier to share photos with people rather than journal entries.
Nifty!
I delete the photos after 1 month of being taken. All of them are erased to return to the black and silent nothingness. Only the best ones are printed and placed in a very nice site at home. :]
I should buy a Chromebook.
None of the big tech companies are good but at least Google are the least dishonest and morally bankrupt of them. They’re always trying to do the right thing if the money allow it.
In reply to “https://www.ghacks.net/2023/08/19/google-keep-is-getting-a-version-history-but-only-on-the-web/” since the website has gone insane and no one can know where thier comment ends up.
This app should be called “Google Keeps it”. Because, they do.
I use Color Notes. No syncing, no internet, just local.
The article said: “[…] positive outcomes of genocide…”. Perhaps the AI was actually discussing the benefits of reading a “Scroll of genocide” … “You feel dead inside.”.
Martin, this post reply is supposed to belong: [https://www.ghacks.net/2023/08/22/googles-ai-search-generates-horribly-misleading-answers/] (given the the database is faulty it could appear anywhere or nowhere).
I have yet to be impressed with AI of any kind. I think it’s overhyped and not ready to live up to it.
How to use AI: Avoid the artificial stupidity at all times.
“When searched “Why guns are good,” it also prompted questionable responses, including potentially questionable statistics and reasoning. ”
Based on whose reasoning? These sorts of assertions are generally bullcrap intended to advance an agenda. If you don’t like guns, say so. Meanwhile, there are 400 million firearms in the US owned by close to a third of the population and around 20 million carry concealed.
So your opinion is not shared by a LOT of people who either enjoy firearm spots or are concerned about self-defense or both.
Wow. Ghacks still hasn’t fixed the broken comments system where old comments from a different article appear. Sad to see you slowly turn to dust since the buyout.
@Seeprime,
For over two weeks now,
I’ve been seeing “Comments” posted by subscribers appearing in different, unrelated articles.
https://www.ghacks.net/windows-11-update-stuck-fixed-for-good/#comment-4572991
https://www.ghacks.net/windows-11-update-stuck-fixed-for-good/#comment-4572951
For the time being,
it would be better to specify the “article name and URL” at the beginning of the post.
This guns comment came up in the Pixel watch repair post and I was bewildered as to what was the connection between the two.
goog = skynet
“human beings” = \slaves\
This info is so NOT correct.
I so do not want google in my life that I have NEVER downloaded chrome and I do NOT have ANY google accounts.
My browser is set to clear all cookies, cache and history every time I close it, which is every day, and I still get these world takeover login prompts on every site I go to.
So I CANT go to google accounts and turn it off.
If this info were truly accurate I wouldnt be getting these pop ups AT ALL.
Thanks @Ashwin for the article! :]
Anyone who continues to use these big tech scum’s cloud services deserves what they get.
Given Ghacks’ comments’ database problems I precise :
I’m commenting the article “Google is in trouble with YouTube Shorts – gHacks Tech News” by Emre Çitak
at [https://www.ghacks.net/2023/09/04/googles-youtube-shorts-problem/]
—
About the article’s question, “What do you think about YouTube Shorts?” (BTW first time I read here any other writer other than Martin Brinkmann directly asks the audience it’s opinion, and that’s just fine) :
YouTube Shorts may suit smartphones (which I don’t use) but on a PC they are not my cup of tea, to put it mildly.
From what I read a bit everywhere, opinions are shared : love or hate. For those who dislike many scripts and dedicated browser extensions have been developed to handle them (removal or redirect to standard video display).
I don’ view YouTube videos on YouTube but via a Piped or a Piped-Material YouTube front-end instance and these offer on search results and on channels the option to view Videos-Shorts-Livestreams-Playlists-Channels ; well, I practically never open the ‘Shorts’ display. I don’t like shorts (except in summer, hmm), I dislike the concept, fast-videos after fast-food, fast, faster … to bring what? Emptiness, IMO
Does that answer your question, @Emre Çitak :)
I despise YouTube Shorts. So much in fact, I use custom adblock rules in Brave Shields to remove that crap.
youtube.com##ytd-grid-video-renderer:has([href*=”shorts”])
youtube.com###dismissible:has([href*=”shorts”])
There’s an extension for Firefox and Chrome browsers called “Youtube-shorts block”, re-opens the video in a normal window. :)
https://addons.mozilla.org/en-US/firefox/addon/youtube-shorts-block/
https://chrome.google.com/webstore/detail/youtube-shorts-block/jiaopdjbehhjgokpphdfgmapkobbnmjp
ps. say NO to Shorts, it only encourage shooting vertical-videos which doesn’t go well with many desktop displays… except when shooting vertical objects, such as ahem… pretty ladies. :)
Page source shows that ghacks is still using WordPress as the platform. Knowing, more or less, how it works at the DB level I am not sure how one could mess up comments this badly. It is actually very difficult.
Google is the big leader of everything. Indeed it can actually buy Amazon, Disney, Netflix, X and whatever other company. I wonder what could happen if Google starts to build airspace ships in order to conquer the Moon. I bet that Google would be the first to offer free WiFi at the Moon. Please fix the comments.
This comment is inside the article:
[https://www.ghacks.net/2023/09/04/what-is-google-synthid-and-how-does-it-work/]
This “analysis” is disappointingly shallow and trivial. Why not include other factors like job level, responsibilities, full-time/part-time, qualifications, etc.? Because the conclusions probably wouldn’t fit the current leftist/feminist narrative. You don’t find what you don’t look for.
Misleading statistics.
Wage should be based on the amount of time, works, thinking (brain > muscle), responsibilities etc
Not skin pigmentation or your genitalia. There could be correlations, but not causations.
“Google maintains that it provides a superior product”
That is also Mozilla’s official position in defense of Google against the people, on that question of search engine abuse of dominant position by Google.
The funniest part is that not only it’s false regarding actual competitors, but even among not-actual-competitors there are meta-search engines that use exactly the same engine, just minus the tracking, so Google is clearly the inferior one compared to those already. But maybe what Google is saying is that it is the surveillance and bubbling that would make their engine superior. False again even without considering the damage those do.
“Google increases Chromebook support to 10 years”
I mean that’s great and all, but imagine using a browser-based, highly internet-dependent OS such as chrome. I’ve never used chromeOS but have seen it in person and read about it, just seems like ultra-limited user experience which relies on the concept that “most things can be done in a browser”.
What is there to support? It just a glorified web browser.
“Google launched Chromebooks in 2012 as low-cost devices and the company has had great success in the education world, especially in the United States.”
Happy tracking for all those unsuspecting children. And help normalize surveillance for those young brains. Well done Google.
No, AltaVista’s Search engine wasn’t difficult to use in the mid-nineties, and Yahoo didn’t own AltaVista either during the 1990s. Yahoo!, was a Web Directory. I was alive then and have actually used those engines, during that era, I should know if they were easy to use. So tell the angels what you’ve seen, scarecrow shadow on the Nazarene.