The Speechify text-to-speech app permits its over 50 million customers worldwide to transform any textual content, together with paperwork, articles, PDFs, and pictures into audio, with over 200 voices to select from. Now, the corporate is delving into a brand new sort of audio: AI-generated podcasts.
Beginning at this time, Speechify customers will have the ability to flip any content material right into a “lecture-style” podcast. They’re going to additionally get entry to a slew of different AI-powered options, together with updates to AI Summaries and AI-generated quizzes and a brand new AI recap characteristic, and maybe the most important, AI podcasts.
Maintain studying to learn how the brand new AI podcast characteristic works and the way it compares to viral instruments corresponding to Google’s NotebookLM Audio Overviews AI podcast characteristic or ElevenLabs’ AI podcast possibility.
Who’s Speechify for?
Speechify is geared towards individuals who both have a studying obstacle or simply choose voice content material. The brand new releases are positioned as examine aids and back-to-school instruments for college kids.
How does the AI-powered podcast work?
On the Speechify platform, you might have at all times been in a position to add your personal supplies for it to learn them out loud for you. Now, these supplies can be utilized to create a podcast. One characteristic that differentiates it from rivals like NotebookLM is that on the Speechify platform, you possibly can select the type of “present” that you simply wish to hearken to, with choices together with a regular podcast, debate, late-night with notes of comedy, and lecture.
As an alternative of importing your personal textual content, you too can enter a textual content immediate of a subject you are keen on, and Speechify will generate a podcast about it for you, one other characteristic that units it other than NotebookLM. Speechify doesn’t delineate what supply materials it makes use of to generate a podcast from scratch, however informed me that this characteristic is powered by its proprietary massive language mannequin (LLM), which is comprised of open and closed supply fashions. The corporate says it runs fixed evaluations to make it higher, however because it pulls from the net and the podcast does not embrace citations, it is sensible to confirm any statements or claims, particularly in the event you’re conducting analysis for varsity.
The way it labored for me (and how one can entry)
Utilizing the characteristic was intuitive. Step one was to click on on “AI Podcast.” When you do, you’re offered with the 4 choices and an outline of every. For instance, the podcast mode is described as “impartial tone, two audio system,” the late-night present is described as “notes of comedy, three audio system,” the talk is described as “argumentative, three audio system,” and the lecture is described as “instructional, two audio system.”
When you click on in your mode of alternative, you possibly can both click on on the “describe with textual content” or “choose file” choice to create the podcast. For the sake of this experiment, I uploaded a PDF of a ZDNET article I wrote. After you add a file, you possibly can attempt the normal text-to-audio characteristic, wherein a voice reads the content material to you phrase for phrase (ElevenLabs gives this characteristic as effectively). That is fairly neat in itself; I had Gwyneth Paltrow learn my story to me.
I selected the AI podcast possibility within the podcast mode, which took one minute and 25 seconds to create. On this planet of near-instantaneous AI, I used to be shocked that it took a bit greater than a minute, however all issues thought-about, a minute continues to be fairly fast –NotebookLM’s course of takes a couple of minutes.
The ensuing podcast was nice. It was about 5 minutes (half as brief as what NotebookLM at the moment gives), correct to what I wrote, and appeared like a human dialog. A singular characteristic I discovered useful was a transcript of the dialog which you can simply observe together with because the phrases are highlighted as they’re learn, like when singing karaoke. Different useful options allow you to management the velocity and simply revisit the unique textual content, which is displayed on a banner on the backside of the app web page.
I then tried it once more with the late-night present and debate modes, because the nuances concerned in carrying these out sounded probably the most intriguing to me. The late-night present was surprisingly higher than I anticipated, as I used to be nervous the three-person chat could be overwhelming. As an alternative, there was respectable banter and hand-off between the three protagonists that made it way more gratifying to hearken to than the usual. The humor added vivid analogies and metaphors that labored rather well to breathe stay into the content material with out altering its that means — doubtless particularly useful for denser matters.
The controversy mode gave me one speaker’s intonation that I couldn’t bear to hearken to due to the chipper, excessive intonations that have been unnatural to what was being stated — however this is not a deal breaker to me since you might have the choice to choose which mode you would favor, and somebody could also be a fan of the overly cheery.
Speechify Premium customers can attempt the AI Podcast characteristic on iOS at this time; net and Android are nonetheless on the way in which, although Speechify didn’t specify a launch date.
Different options
Speechify has already been utilizing AI to gas different options on the platform, together with producing summaries and quizzes, earlier than transferring into podcasts. These options additionally obtained an improve that enables customers to create summaries and quizzes from particular chapters inside their studying quite than the whole textual content.
AI Recap is a brand new AI characteristic that enables customers to click on “recap” to get a abstract of what they simply learn, just like the refresher you see earlier than watching the subsequent episode of a TV present.
Speechify’s Chrome Extension obtained some updates, together with the addition of AI Summaries. This lets customers summarize any textual content instantly of their browser’s aspect panel. Customers can even hone in on a part of the textual content by utilizing the “ask AI” characteristic. Speechify shares its Chrome Extension was additionally upgraded to be sooner, loading virtually immediately with out slowing down looking.
Some non-AI additions embrace the flexibility to take notes, spotlight textual content, bookmark readings, search in paperwork, share information, and improve voices that sound extra conversational. Speechify can also be enabling a filtering expertise that makes it simpler for customers to seek out simply the precise voice for them.
Lastly, within the back-to-school spirit, Speechify additionally added over 70,000 free books to Speechify, together with traditional literature, OpenStax’s catalog of free textbooks, and 130,000 textbooks from different publishers corresponding to Taylor & Francis and O’Reilly that customers can entry instantly within the platform.
