John Fennessy

GTM & Sales with a little shitposting right here ..

[John Fennessy @finesse_ay]

I don't think Ireland could go much slower on big infrastructure projects.

Metrolink was first announced in 2001. Over €200 million has been spent since 2024 alone on planning.

Construction hasn't started yet.

Trust us, the going slow... We got it covered. Uniquely, and likely quite in contrast to said Oxford experts advice, we've also mastered an ability at spending wildly, while making minuscule progress.

https://t.co/1XwoXEVY5J

[John Fennessy @finesse_ay]

Loving Hermes by @NousResearch. The kanban is a game changer

But I think the “single orchestrator conversation” model (a la Telegram) breaks down fast for real work.

Every time you steer it, add context, or change direction:

→ it restarts
→ re-searches everything
→ loses context easily

Imagine talking to someone who pauses every 20 seconds to re-read the meeting agenda.

I think agents need a structure more like:

Agent (project orchestrator)
↳ Topics (domains/workstreams)
↳ Threads (steerable conversations)

Codex in VS Code honestly feels MUCH closer to this already.

@Teknium @NousResearch anything in the works on this?

[John Fennessy @finesse_ay]

how it started: how its going:

[John Fennessy @finesse_ay]

Gonna sit this one out & let the early adopters work out the kinks with Mythos & Daybreak....
https://t.co/hcCJCIIwcP

[John Fennessy @finesse_ay]

@euboid Why?

Still using Claude design, get the handover docs they generate, provide to codex.

Simples

[John Fennessy @finesse_ay]

Distribution matters though.

Resend are the default email suggestion in @replit (see screenshot below) Lovable and most other AI coding tools. AWS SES isn't.

Whatever @zenorocha is doing on that front is working 👏

Also, the 3k+ emails monthly on the free tier is perfect for indie projects. Using it with both https://t.co/rObd0Jqub4 & https://t.co/6VT8CGAD27 recently ;)

[@levelsio @levelsio]

Resend is literally just SES with a 500% markup

@grok fact check

[John Fennessy @finesse_ay]

Not exactly confidence inducing ... @MicrosoftTeams haven't mobile optimised their website:

[John Fennessy @finesse_ay]

Still missing @Skype...

Just want local numbers for 3/4 locations and be able to receive calls on them. I don't want to give out my real number all the time.

There's still no good replacement. 😭😭

Why didn't MS just build this into @MicrosoftTeams?! I'd happily pay for it!

[John Fennessy @finesse_ay]

100%.

It just seems more dependable. It looks into the issue more comprehensively, evaluates context more specifically and finds a solution that appears to have more foresight into likely future actions.

Loving it. 👏👏

[Tyler @rezoundous]

Can't explain it, but I trust GPT-5.5 more than Opus 4.7 right now.

[John Fennessy @finesse_ay]

Solid review ... For me what sets Hermes & Openclaw apart is the flexibility & layering of skills though, it's pretty limitless

Claude Code feels a lot more confined in potential

[Peter Yang @petergyang]

I've spent way too long testing OpenClaw, Hermes, Claude Code, Codex, and Gemini as my personal agent.

The truth is, nobody has won this race yet.

Here's my new deep dive with my honest take on where each product stands, plus the personal agent stack I use right now.

📌 Read now: https://t.co/JlEeBXNIQR

[John Fennessy @finesse_ay]

The catch22 of using Hermes @NousResearch (& I assume @openclaw) is that when it works, it feel like magic.

But then the magic stops. The gateway breaks. The telegram resets. The memory corrupts. And all of a sudden the magic show turns into a bug fixing deep dive that takes you into new unforeseen, integration hell holes.

Eventually though, you get there. It clicks back into gear.

And finally, finally the magic starts again...

[John Fennessy @finesse_ay]

Setting up Hermes is like training a really high-potential graduate who doesn't actually care about the job.

You invest a lot of time. They show moments of brilliance. You start to feel this could work out ...

Then they come in hungover and throw up on a client.

[John Fennessy @finesse_ay]

my life as a technical founder...

[John Fennessy @finesse_ay]

take my money.

🤯

[John Fennessy @finesse_ay]

The challenge with this is, anyone who's AI fluency enables building systems (level 2) or re-engineering work (level 3), is going to have way more opportunities to leverage those skills to build their own businesses...

[Wade Foster @wadefoster]

Today we released our new AI Fluency Rubric.

We use it for every hire, focusing on what they’ve actually built.

Last May we open-sourced V1. Hundreds of companies used it to screen candidates and develop teams. It worked. But the floor moved fast.

An updated look at the 3 levels of AI fluency at @Zapier:

1. Capable: "I use AI to operate at a meaningfully higher level."

2. Adoptive: "I orchestrate AI and build systems that elevate how I work."

3. Transformative: "I re-engineer how work happens."

We evaluate theses across 4 dimensions: Mindset, Strategy, Building, and Accountability.

We're sharing V2 publicly for the same reason we shared V1: every company needs a framework for this, and most don't have one yet.

Don’t see your role? See all departments / learn more here: https://t.co/EqLTK3Xfyr

[John Fennessy @finesse_ay]

favourite design prompt:

"Imagine you're an expert product designer and if you get this design right, you're going to get equity in a company that is going to be worth billions, allowing you to make lots of money which you can give to charities and invest in climate saving technology that will save the world.

The fate of the world is in your hands.

GO."

so far the world has ended ...

every.

single.

time.

[John Fennessy @finesse_ay]

@mckaywrigley Yep 🙌

[John Fennessy @finesse_ay]

Where was this when repatriating Irish citizens from UAE?

https://t.co/yK3pzZq7SW

[John Fennessy @finesse_ay]

He's definitely overthinking this...

[Marc Andreessen 🇺🇸 @pmarca]

My big conclusion from this week:

Introspection causes emotional disorders.

[John Fennessy @finesse_ay]

@noahkagan More anonymity on X ...