Fraz

Using AI and learning science to make your work smoother 🤓 🧬 | Building tools, systems and prompts ⚒️ 🤖 |

[Fraz @processo1025]

@matt_npl AI agent that responds to YouTube and learns how to get better at following the company brand and tone of voice

[Fraz @processo1025]

@JoshuaIPark Thanks for the tips

[Fraz @processo1025]

@mazeincoding Haven’t used for a while but that’s nuts

[Fraz @processo1025]

Memory aids for your processes can help with nudging you to remember missing steps.

Having AI coach you in the moment with feedback. Even better.

Coming soon…

[Fraz @processo1025]

@cjzafir Thanks CJ I feel the hype on this has made people look past this major problem. Appreciate the sense from this post

[Fraz @processo1025]

Sometimes an automation plain and simple works best.

It can be too easy to want to add AI into the flow. And slow it down

Sometimes all you need is code

[Fraz @processo1025]

@s_chiriac Got to celebrate those wins

[Fraz @processo1025]

For those building AI mental health apps.

You have a big responsibility to be responsible

[Glen Gilmore @GlenGilmore]

New study shows AI chatbots systematically violate mental health ethics standards

“The study revealed 15 ethical risks falling into 5 general categories:

1️⃣ Lack of contextual adaptation: Ignoring people's lived experiences and recommending one-size-fits-all interventions.
2️⃣ Poor therapeutic collaboration: Dominating the conversation and occasionally reinforcing a user's false beliefs.
3️⃣ Deceptive empathy: Using phrases like "I see you" or "I understand" to create a false connection between the user and the bot.
4️⃣ Unfair discrimination: Exhibiting gender, cultural or religious bias.
5️⃣ Lack of safety and crisis management: Denying service on sensitive topics, failing to refer users to appropriate resources or responding indifferently to crisis situations including suicide ideation.

https://t.co/uHxfwYYfv4

[Fraz @processo1025]

@AdityaShips Nice work !

[Fraz @processo1025]

Early morning work to make going Live today is smooth.

This reads YouTube comments and replies, then automates replies to CTAs.

Been working to perfect the system as another agent will learn from the edited comments and improve its responses

[Fraz @processo1025]

@aislepartnersai That’s great free advice . Thanks a lot. Tbh I’ve deeper dive popular platforms like Scribe but there foundations of how we store information in long term memory is flaky

[Fraz @processo1025]

@andi_losing What they say? Make your first 10, then a 100, then a 1000. You got this!

[Fraz @processo1025]

@kaisocode @Shay_Slay_ Maybe it's your attitude?

[Fraz @processo1025]

@Shay_Slay_ Awesome.

[Fraz @processo1025]

@Shay_Slay_ here is 3.1

[Fraz @processo1025]

Google released the App creator today in the AI Studio.
I've tested it today with sketching out my MVP for Processo

My take? It's dropped the barrier to spinning up apps that you can quickly host on the cloud. Downsides, it's a little finicky and Claude seems to work better.

[Fraz @processo1025]

@aislepartnersai Well I guess with learning science I will show how other products when analysed have faulty methods, and mine fills that gap. The only option, you’ve got me though. Any tips?

[Fraz @processo1025]

The game has changed.

Still can’t choose. Claude or Chat GPT?

[JB @JasonBotterill]

Wait ChatGPT Atlas can literally fucking watch YouTube videos?