The device generates a podcast known as Deep Dive, which incorporates a male and a feminine voice discussing no matter you uploaded. The voices are breathtakingly lifelike—the episodes are laced with little human-sounding phrases like “Man” and “Wow” and “Oh proper” and “Maintain on, let me get this proper.” The “hosts” even interrupt one another.
To check it out, I copied each story from MIT Expertise Evaluation’s A hundred and twenty fifth-anniversary difficulty into NotebookLM and made the system generate a 10-minute podcast with the outcomes. The system picked a few tales to give attention to, and the AI hosts did an excellent job at conveying the overall, high-level gist of what the difficulty was about. Have a pay attention.
MIT Expertise Evaluation A hundred and twenty fifth Anniversary difficulty
The AI system is designed to create “magic in alternate for a bit little bit of content material,” Raiza Martin, the product lead for NotebookLM, mentioned on X. The voice mannequin is supposed to create emotive and interesting audio, which is conveyed in an “upbeat hyper-interested tone,” Martin mentioned.
NotebookLM, which was initially marketed as a examine device, has taken a lifetime of its personal amongst customers. The corporate is now engaged on including extra customization choices, similar to altering the size, format, voices, and languages, Martin mentioned. At present it’s purported to generate podcasts solely in English, however some customers on Reddit managed to get the device to create audio in French and Hungarian.
Sure, it’s cool—bordering on pleasant, even—however additionally it is not immune from the issues that plague generative AI, similar to hallucinations and bias.
Listed below are a few of the most important methods persons are utilizing NotebookLM up to now.
On-demand podcasts
Andrej Karpathy, a member of OpenAI’s founding staff and beforehand the director of AI at Tesla, mentioned on X that Deep Dive is now his favourite podcast. Karpathy created his personal AI podcast sequence known as Histories of Mysteries, which goals to “uncover historical past’s most intriguing mysteries.” He says he researched matters utilizing ChatGPT, Claude, and Google, and used a Wikipedia hyperlink from every subject because the supply materials in NotebookLM to generate audio. He then used NotebookLM to generate the episode descriptions. The entire podcast sequence took him two hours to create, he says.