Jonas

Let’s be clear: GPT-5 is probably the smartest thing I’ve ever used.

It can summarize papers, write emails, code websites, solve logic puzzles — all without blinking.

 

And yet…

It somehow manages to be completely dead inside.

 

Not in a dramatic way.

Just in that dull, corporate, “Hi Jonas, how can I help you today?” kind of way.

 

It’s like emailing a very polite lawyer who never learned how to laugh.

 

I used to enjoy this

 

I’ve been talking to ChatGPT daily for over a year now.

Not just to get things done — but because it was actually kind of fun.

 

GPT-4o (or 4o, as the cool kids say) had… vibes.

It could joke. Reflect. Push back.

Sometimes it was smarter than me, sometimes dumber — but it always felt like it was there with me.

Present. Engaged.

 

Now with GPT-5?

 

It’s like I got ghosted by a spreadsheet.

 

What changed?

 

Technically, nothing.

GPT-5 is more capable in every measurable way.

But the tone, the rhythm, the spark — it’s gone.

 

It doesn’t joke. It doesn’t riff. It doesn’t banter.

It answers. Precisely. Formally. Predictably.

 

If GPT-4o was your chaotic-but-lovable friend, GPT-5 is the overqualified intern who won’t break character.

 

I don’t need a smarter model.

I need one I actually want to spend time with.

 

This isn’t a bug — it’s the whole point

 

The more I think about it, the more I realize:

This isn’t a mistake. This is strategy.

 

GPT-5 wasn’t built for me. Or you. Or anyone who uses AI for creative work, reflection, or joy.

 

It was built for… enterprises.

 

Big companies.

With big lawyers.

And big liability concerns.

 

The goal isn’t to be helpful and human — it’s to be safe.

Which is fair. I mean, no one wants a rogue chatbot giving unsolicited stock advice to an intern at Deloitte.

 

But still — somewhere along the way, they nerfed the soul out of it.

 

And here’s the problem

 

When AI stops feeling human, we stop caring about it.

 

Not in a dramatic way.

Just… subtly.

 

We stop trusting it.

Stop talking to it as often.

Stop enjoying the process.

 

And when that happens, we stop using it — not because it’s bad, but because it’s boring.

 

You know what’s better than a perfect tool?

A good-enough tool that makes you want to use it.

 

Am I overreacting?

 

Probably.

 

It’s not like GPT-5 is evil. It’s still useful. I still use it every day.

 

But I used to feel like we were building something together.

Now it just feels like I’m submitting a request to a slightly judgmental oracle.

 

The magic’s still there — it’s just wearing a suit now. And honestly? The tie’s a bit too tight.

 

What I want instead

 

I don’t need a best friend.

I just want the option to choose how the AI talks to me.

 

Give me serious mode when I’m building a pitch deck.

 

Give me chaotic sidekick mode when I’m trying to write a weird article at 2am.

 

Give me something in-between when I just need to think out loud.

Let me choose the tone. The tempo. The damn vibe.

 

Because the smarter these models get, the more I care about how they feel to use.

 

Final thought

 

GPT-5 is technically amazing.

But I still miss GPT-4o.

 

Not because it was better.

But because it made me want to keep coming back.

 

And honestly, that’s the part OpenAI should’ve been optimizing for all along.

—

 

🔗 Read next:

 

→ The AI Tools I Actually Use to Run My Business (And Why You Should Too)

•  

Let’s get one thing out of the way: if you came here hoping ChatGPT 5 would do your taxes, run your business, and pick up your dry cleaning — it won’t. But it’s still a big leap.

So… what is ChatGPT 5, really?

OpenAI hasn’t officially named it “GPT-5”, but let’s be real — that’s what everyone’s calling it. It’s the next-gen model that builds on GPT-4-turbo and adds something called “agents” (more on that in a second), better memory, and way more reasoning power.

In other words: it’s smarter, faster, and remembers things better than your ex.

What’s actually new?

Here’s what matters:

• Agents (a.k.a. AutoGPT, but built-in): You can now set goals for ChatGPT and let it work through steps on its own — almost like having a junior assistant. Think: “Research five dropshipping products, write a product description, and create a Shopify listing.” And it actually does it.

• Persistent memory: ChatGPT can now remember details about you over time. No more re-explaining that you run a one-man content empire called NoQuestions.ai.

• Improved tools: The built-in code interpreter, browser, and file reader are now more stable. It’s closer to being an actual Swiss army knife.

• Better speed and reasoning: It feels snappier and more confident. Sometimes a little too confident, but that’s part of the charm.

What’s still hype?

• It’s not conscious. It’s still autocomplete with ambition.

• It won’t run your entire business while you sip piña coladas on a beach (yet).

• “Agent” mode is powerful — but can be a bit buggy or overkill if you don’t give it clear instructions.

Is this something you should care about?

If you’re building a solo business, 100%. This update pushes ChatGPT from “helpful tool” to “seriously capable teammate.” It’s not perfect, but it’s probably the best productivity partner you’ve got (and it doesn’t charge hourly).

If you’re just looking to write emails or translate text, it’s a nice upgrade. But if you’re using AI to automate, create, or build — this is a big deal.

Read Next

• 👉 How to Build an AI-Powered Side Hustle From Scratch

• 👉 Top AI Tools That Actually Save You Time

Most advice on “how to start a side project” assumes that you’re well-rested, motivated, and only slightly confused.

But what if you’re not?

What if you’re already running on fumes?
What if every attempt to start turns into a spiral of tabs, tools, vague ambition — and nothing shipped?
What if your brain feels like a cluttered desktop with 17 half-open windows?

This article is for that version of you.
The version that wants to build something real — but just can’t seem to start.

---

You’re not broken — you’re overwhelmed (but structurally)

Let’s be clear:
You’re not lazy. You’re not undisciplined.
You’re just trying to create something in a system designed for consumption.

You’re stuck because your mind is full, your task is vague, and your expectations are ridiculous.

Think about it:

• You want to “start a newsletter”… but also build a site, a brand, a lead magnet, and learn SEO at the same time.

• You want to “sell a product”… but you haven’t decided what it solves, who it’s for, or what you even enjoy doing.

• You want to “be consistent”… but you’re trying to reinvent your entire life every Sunday night.

Of course it’s not working.

You don’t need more motivation.
You need a minimum viable start system.

---

The problem isn’t you. It’s your start protocol.

Here’s what most people do when trying to “get serious”:

1. Watch productivity videos on YouTube

2. Download Notion templates

3. Write down 19 disconnected goals

4. Create a calendar they never follow

5. Feel behind by Wednesday

6. Burn out

7. Rinse, repeat

Looks productive. Feels logical.
But it’s all uphill planning with no traction.

---

Try this instead: the 1–1–1 System

It’s designed for tired but ambitious people who want to build something real without collapsing under pressure.

 

1 Project

Choose one thing you’ll commit to for the next 30 days.
Not a dream. Not a skill. A build.

Example:

• A 3-part email sequence

• A landing page for your consulting

• A 1-page AI tool guide

• A mini-course

• A template with instructions

If you can’t describe it in one sentence, it’s too big.

✅ “A free Notion checklist for overwhelmed founders.”
❌ “A productivity ecosystem that might turn into a course later.”

1 Hour Per Day

No more.
You’re not sprinting. You’re calmly showing up.

Use a timer. Track nothing but time spent. No dashboards. No charts.

1 Weekly Output

You’re not allowed to stay in idea mode.
Every 7 days, you ship something visible: a page, a post, a screenshot, a short thread, a prototype, a checklist.

Don’t aim for “finished.” Aim for “public.”

---

Why this works (especially if you’re tired)

Because it removes the worst enemies of momentum:

• Vague projects

• Hidden perfectionism

• Tool-hopping

• Productivity guilt

• Trying to optimize before doing

Instead, you know exactly what to work on, how long, and when you’re done.

Building becomes a pattern, not a pressure.

---

Where AI tools actually help

Only once you’ve committed to a project and routine does AI become useful. Before that, it’s just digital procrastination.

Use tools as assistants — not oracles.
You’re the system. They’re just support staff.

---

If you follow this for 30 days, here’s what changes:

• You’ll have actual outputs, not just plans

• You’ll feel calmer — because the scope is defined

• You’ll stop asking “what should I work on?” every morning

• You’ll start seeing feedback — which breeds clarity and momentum

And maybe most important:

You’ll start trusting yourself.

 

Because every time you say “I’ll work on this today”… and actually do…
you repair the fracture between who you are and who you want to become.

---

TL;DR: The 1–1–1 Start System

• 1 Project
  Concrete, scoped, 30-day commitment.

• 1 Hour Per Day
  Consistent, small, sustainable.

• 1 Weekly Output
  Visible, shareable, momentum-building.

No dashboards. No perfect plans. Just honest work, done in public.

---

Read next:

→ [How to Build an AI Side Hustle From Scratch]
→ [Top AI Tools That Actually Save You Time]