Bard & ChatGPT, Music, & Psychiatrists

On today’s 2-minute edition of the AI Alchemist, we will discuss all they hype around Google, medical use-cases, text-to-music, and more.

Many are saying its latest advancements with Bard outshadow those of ChatGPT. Even all the Twitter threads going “ChatGPT is DEAD. Here are 9 things you can do with Bard….” 🙂 But…we think this may not be the case!

Here’s why: ChatGPT, much like the iPhone, has done the most difficult thing a company can do. Much like the iPhone, ChatGPT captured the awe, imagination, and Zeitgeist of hundreds of millions of people around the world. It garnered a strong network effect, created a category, and even injected itself into culture.

Every other reel on Instagram has something to do with ChatGPT. South Park episodes even have it. Hundreds of startups have the name GPT in their domain name…RizzGPT anybody?

Anyways, the point remains: ChatGPT has defined the category, and is forever inscribed in human-AI consciousness.

Whoa…I gotta tone it down, don’t want to get all philosophical!

Sure, Bard has things it could be better at, simply because of access to Google’s reach, data, and scope- such as workflow tools. BUT, in this category, it will forever be the laggard, the one playing catch-up from a brand perspective.

But, the war will continue….

__________________________________________________________________________________

Your AI Psychiatrist?

Hot off the press, Palo-Alto clinic is releasing an AI-first psychiatrist. Now, this is a very delicate avenue. This is because medical information is very sensitive data, and cannot be handled lightly.

In a new Twitter space by one of the founders developing this technology, Vikas Kedia, emphasized the need for AI startups dealing with this sensitive information to need to develop in-house systems to ensure the integrity of the information and data. Simply slapping together an API doesn’t seem to cut it- gotta feed the models with proprietary data, and ensure the privacy of user data you collect.

Either this sounds very difficult, Dr.GuptaAI’s been doing it right!

Anyway…

Music LM

Remember how we talked about music AI-genereated music, and the potential copyright issues that it would face?

Well…Google just released MusicLM, a language model that helps people generate music!

One user from Twitter (Rob Lennon) was able to produce exquisite sounds with it. Here’s some of the prompts he used:

• epic star wars cinematic soundtrack and strings and brass for fight scene

• Irish folk music with fiddle and bagpipes

• notorious 90s hiphop to make you feel like a baller

• lo-fi music to imagine the potential of the future to

• drum core beats on the field during half-time

• tribal acid jazz that creates an eerie feeling

Hey, not bad, Google. You may be onto something here…. (unlike Google +!)

Big News in AI:

Elon’s meeting with CNBC:

In this interview, Elon said “I’m the reason that OpenAI exists!”

This is funny, and true. See, Elon invested $50 million to make OpenAI a open source, non-profit company.

It is now, however, a closed source, capped-profit company.

Fair criticism. However, Sam Altman stated this shift was done in order to secure enough funding for AI development.

Hey, don’t hate the player….or the game!

Anyways, tech advancements, drama….AI’s got it all this week.

In the next edition, we’ll talk about AI in finance, academia, and more business ideas….

Peace out!