The 💯 Digest: GPT-5 dropped and everyone's confused

👋 Hey, it's Max

It's been a ridiculous week. Claude's flexing, Google's teasing, Genie 3 dropped and oh yeah, OpenAI released GPT-5.

The AI world just hit fast-forward handing developers, creators, and businesses more power than they know what to do with.

So I’m sure you’re as overwhelmed as we are and could use:

• A recap of why GPT-5 reviews are all over the place

• What Google's Gmail integration is about to do to your workflow next week

• 2 tricks to make GPT-5 actually useful by Monday

• Plus useful links (and memes) to drop in your Slack and look smart

If your feed is anything like mine, it must be split in two:

Some are saying the model is really good; Theo from T3 Chat says he’s horrified because he didn’t know it could get this good, Matt Shumer is calling GPT-5 the best coding model in the world, and Claire Vo felt like GPT-5 was an engineer, built by engineers, for engineers.

Meanwhile, Pieter Levels is 🤬 because he can’t switch to 4o and says this is affecting his productivity. Thousands are flooding Reddit with complaints; calling the new model "horrible," "overly verbose," and claiming it's actually performing worse than GPT-4o on basic tasks. Some users report getting completely wrong outputs or suspecting they're being served the wrong model entirely.

Our own team is split too. Our Growth lead G thinks it's great for coding but still outputs predictable AI aesthetics. While Ciara (our AI Education Lead) said GPT-5 stripped out all of 4o's personality and shared this tweet pointing out deeper concerns about consistency.

Ethan Mollick put it perfectly: "You are likely going to see a lot of very varied results posted online from GPT-5 because it is actually multiple models, some of which are very good and some of which are meh. Since the underlying model selection isn't transparent, expect confusion."

Read Mollick's full breakdown

Here's the pattern I've seen with every major AI release: massive hype on day one, then crickets by week two. Same thing happened with GPT-4, everyone poked around, made memes, then quietly went back to their old workflows.

Don't be that person. Here's how to actually leverage this on Monday 👇

Practical Tip #1:

Practical Tip #2: OpenAI's Secret Sauce 

OpenAI just dropped a massive prompting guide. Here are the gold nuggets for non-technical folks:

• Break complex asks into sequential steps (GPT-5 handles multi-stage reasoning better)

• Use "Let's work through this systematically" to trigger its analytical mode

• For creative work: Start broad, then narrow with follow-ups (it maintains context better than 4o)

• The sweet spot: 2-3 paragraph prompts with clear context, not one-liners

Full OpenAI Prompting Guide

GPT-5 is powerful and polarizing, but the magic is less in the tool and more in YOU knowing what problem you want to solve with it.

If you only use it when you're stuck, it'll be a nice sidekick. But if you deliberately redesign parts of your job around it, it changes everything. Proactive vs reactive use of AI is what seperates the top 1% of users and the rest.

So pick one tip from above, try it this week, flex on your team, and let me know how it goes.

Until next time (drop me a reply if you have any Q’s 😊),

Max 👋

P.s Want to make your team & company AI-first? Let us help here.