Today at a glance:

• GPT-5 launches with mixed reception - I stayed up all night testing it

• Neuroscience research breakthrough - AI accurately predicts brain responses

• ElevenLabs enters music generation - Competition heats up with Suno

• Complete guide: Running OpenAI OSS-20B locally - Your own offline LLM

• GPT-5 video demo - Real use cases with thinking and Pro models

• Music generation for entrepreneurs - Custom tracks for any business need

News: 

GPT-5: Is it worth the hype?

OpenAI officially launched GPT-5, and it's been the biggest topic in AI circles this week.

Some people were underwhelmed, claiming it was all hype, and that OpenAI underdelivered.

As for me, I still think it’s a significant upgrade.

Sure, the GPT flagship model is only slightly better than GPT-4.

But GPT-5 Thinking is pretty good, and Pro is mind-blowing.

I made a YouTube video about it and have done some really cool stuff with it already.

OpenAI will take feedback onboard and do tweaks and updates over the next couple of months, so, if you’re not convinced, just give it some time.

Meta Develops Neural Network to Predict Brain Responses

Meta’s FAIR team just introduced TRIBE, a deep neural network that predicts how human brains respond to stimuli, such as video, audio, and text.

The innovation won first place in the Algonauts 2025 brain modeling competition.

TRIBE analyzes video, audio, and dialogue from movies, and can accurately predict which regions of the viewer’s brain will activate without any brain scanning.

This shows how AI is helping us learn more about how the brain works.

Personally, I can’t wait until we can use AI to create movies from our memories and watch them back when we feel like reliving the past!

ElevenLabs Jumps Into Music Generation

Move over Suno, ElevenLabs has officially entered the chat with their new music generation platform.

As a sound engineer and DJ, this is super exciting for me.

I don’t even want to generate entire songs; I want to describe the beat or sound I'm going for and refine it.

Let’s face it, we don’t want AI to take over the whole music-making process, but it’ll be a great supporting tool for producers and anyone who has an idea for a song.

One Guide :

Run OpenAI OSS-20B Locally: Your own offline AI assistant

OpenAI OSS-20B was last week’s exciting news.

Being able to have an offline, local LLM that you can fine-tune is really cool, and it makes me wonder if I need to buy the latest MacBook Pro.

Or, should get a few GPUs and build a mini data center in my house? 🤣

Ok I’m getting carried away. Here's a guide for you to run your own offline local LLM version of the 20B model:

Step 1: Download LM Studio

Visit LM Studio's website and download the application for your operating system. It's available for Mac, Windows, and Linux.

Step 2: Install and Launch

Install LM Studio like any other application. When you launch it, you'll see a clean interface designed for managing local language models.

Step 3: Find OpenAI OSS-20B

In the model browser, search for "OpenAI OSS-20B" or browse the available models. You'll see different quantized versions (quantization refers to how OpenAI reduces the memory footprint of the models). I recommend starting with the Q4_K_M version for the best balance of performance and quality.

Step 4: Download the Model

Click download and wait. The 20B model is roughly 12-15GB depending on quantization, so make sure you have sufficient storage and a stable internet connection.

Step 5: Configure Your Setup

Once downloaded, you can adjust settings like:

• Context length (I use 4096 tokens for most tasks)

• Temperature (0.7 works well for creative tasks)

• GPU acceleration (if you have a compatible GPU)

Step 6: Test Your Local AI

Start a conversation and test the model with various prompts. You'll notice it's surprisingly capable for local inference, especially for coding, analysis, and creative writing tasks.

Pro Tips:

• The model runs great on 16GB of RAM, but 32GB is even better

• For M2/M3 MacBooks, the unified memory architecture works beautifully

• You can run this completely offline once downloaded

• Perfect for sensitive business data that you don't want to send to external APIs

The experience of having a capable AI running entirely on your machine is liberating, especially for business-sensitive tasks.

One Media:
GPT-5 Explained with Real Demos

When GPT-5 came out on Friday, I didn't sleep all night.

In this video, you'll see how I used it to build fully interactive dashboards and create fun games in minutes. I test GPT-5 Thinking and GPT-5 Pro models side-by-side, revealing GPT-5 in simple, clear terms anyone can understand.

It even makes beats and music. That's awesome!

One Tool:

ElevenLabs Music: Custom Soundtracks for Your Business

As mentioned in the news section, ElevenLabs has launched their music generator.

It's still early days for music generation, but with ElevenLabs getting into it, hopefully we'll see more AI tools that assist with creating tracks, so we don't have to do everything manually.

Why entrepreneurs should care:

You can generate original music for explainer videos, ads, social media posts, and more.

Instead of spending hours searching stock libraries for the right track or paying licensing fees, just create your own in minutes.

Real use cases I've tested:

I created a tidy-up song for my daughter called "Kamila the Lion," and suddenly she loves putting her toys away. After dinner, she asks, "Can you put on the tidy-up song?" and acts like a lion while tidying up.

Now that’s a great use case for entrepreneurs who have kids!

TL;DR:

• GPT-5 launched — Incremental improvements in the flagship model but GPT-5 Thinking and Pro are impressive upgrades

• Meta develops a deep neural network to predict brain responses

• ElevenLabs enters music generation, great for creating custom music to use in your business

• Run OpenAI OSS-20B locally on your Mac for offline AI access

• Discover GPT-5 use cases in my video demo like beat-making and interactive dashboards

• ElevenLabs Music generates custom tracks - I made a tidy-up song for my daughter!