Momentum AI

Momentum AI, through the eyes of founder Craig Turner; trained more than 3,000 orgs on using AI in the real world; creator of Momentum AI Essentials platform.

[Momentum AI @MomentumAIonX]

After training over 3,000 organizations, it's clear: AI training needs to evolve.

Today, we launch Momentum AI Essentials.

Here's why we created it, why it's different, and three ways the platform is now available to individuals and organizations.

https://t.co/w5OQA55p74

[Momentum AI @MomentumAIonX]

Influencers often fail to talk about their timing. They were first. Once everyone is following the influencer's advice, it doesn't stand out anymore.

Find your audience - one that you understand their real-world daily challenges - and do everything in your power to serve them. That's it.

[Momentum AI @MomentumAIonX]

@FrankLuntz The parents' responsibility.

But, Meta can be more responsible with its algorithm.

[Momentum AI @MomentumAIonX]

@maggiemoda Are they?

My experience is that people don't care - as long as you're not trying to disguise that you're using AI.

[Momentum AI @MomentumAIonX]

@strimblez I started that way, but quickly pivoted. Codex's own guardrails are strong, but if you put your own in place on top of them, you're good.

[Momentum AI @MomentumAIonX]

That line of thinking isn't going to work. Here in New York, it really doesn't matter if you use surveillance to catch criminals quickly... Because they just put them back out on the street later that night. You know, so surveillance can catch them again tomorrow. And on Friday. Definitely Saturday. And so on...

[Momentum AI @MomentumAIonX]

@HarperSCarroll If it's making you think less, you're doing it wrong.

For me it's 100% the opposite. The thinking never stops now, because so much is possible that wasn't before.

[Momentum AI @MomentumAIonX]

@coryalthoff If you're building something someone else has already created, you're not taking advantage of the power of AI. You have something you do that makes yours different - a process, a mindset, a level of higher thinking... Customization is king.

[Momentum AI @MomentumAIonX]

@viticci I share the chat with Codex. It's not a one-click solution, but it works great.

[Momentum AI @MomentumAIonX]

I still need to be able to go from computer to computer working on the same projects (Codex/Work). My understanding that capability is available, but it hasn't propagated to my laptop. I'm on the road so much - especially this fall - that I need access to my powerhouse desktop from my laptop. Can you make that happen?

[Momentum AI @MomentumAIonX]

@OpenAIDevs @ChromiumDev @CloudflareDev @ShopifyDevs @vercel @render @Netlify This is awesome! We're in!!!!

[Momentum AI @MomentumAIonX]

That's why we need legislation for spam and phishing safeguards for American businesses. Could you imagine the economic impact if every business person in America didn't have to delete 150 spam e-mail a day?

I always say if I ever run for Congress (I'm not), that would be the first thing I'd go after.

[Momentum AI @MomentumAIonX]

Source note: The durations are calculated from @OpenAI's published launch and change dates. “Experiment happening in public” and the business lessons drawn from these lifecycles are Momentum AI’s interpretation.

Primary sources:

ChatGPT release notes:
https://t.co/Bm41Chdrl5

SearchGPT prototype:
https://t.co/rs806qfsbb

Operator:
https://t.co/gdmynRSNAU
Atlas evolution:

https://t.co/YJRkgpNq2M

Sora discontinuation:
https://t.co/qUN5puSWsY

[Momentum AI @MomentumAIonX]

AI products are not being finished and handed to us.

They are being developed while we use them.

✅Browsing: paused, repaired, relaunched.

✅ SearchGPT: temporary prototype, then ChatGPT Search in 98 days.

✅Operator: standalone research preview, then absorbed into ChatGPT agent in 175 days.

✅Atlas: standalone AI browser, then its capabilities moved into ChatGPT and Codex after 292 days.

✅Sora: research preview, public product, social video app, then discontinued web and app experiences.

Those are not all failures. Some experiments succeeded by becoming part of something larger. Others were repaired, replaced or retired.

The business lesson:

👉Try. Measure. Listen. Adjust.

Do not spend nine months designing a permanent AI plan around technology that may change twice before the committee finishes meeting.

AI is being developed in public. Your organization needs to learn how to adapt with it.

[Momentum AI @MomentumAIonX]

My stuff mostly comes from customer requests. I usually get halfway through a project and realize "this isn't a one-time thing." I have MANY ChatGPT Sites open on my screen, all the time. The good thing is that experimentation takes an hour - not six weeks. "Fail fast" is a great tool for organizing your work.

[Momentum AI @MomentumAIonX]

@sama @thsottiaux @OpenAI @OpenAIDevs @OpenAINewsroom

I thought this was a pretty fascinating timeline. X has only shown the post to 23 people, so I'm tagging you in case you might be interested.

[Momentum AI @MomentumAIonX]

I mapped 63 significant ChatGPT and @OpenAI releases, expansions, experiments, integrations and retirements from November 2022 forward - and the pattern is astonishing.

Not because OpenAI keeps adding features.

Because the products themselves keep changing shape.

- Browsing launched, was pulled back, repaired and returned.

- SearchGPT was introduced as a temporary experiment and became ChatGPT Search.

- Operator lasted 175 days as a standalone product -before being absorbed into ChatGPT agent.

- The Atlas browser (which never even made it to Windows) lasted 292 days before its capabilities began moving into ChatGPT and Codex.

- Sora (video) followed a different path: research preview, public product, social video app and, eventually, discontinued web and app experiences.

... and on and on.

I pulled this timeline together for one reason: to give you a clearer way to think about what AI can actually do for your business now - not what it could do when your organization last evaluated it.

Over the next five days, I’m going to share my thoughts on what this release history reveals:

Monday: The complete timeline - and why I built it

Tuesday: AI products are experiments happening in public

Wednesday: What happens when the product disappears but the capability survives

Thursday: How ChatGPT moved from “ask it a question” to “give it work”

Friday: Why your organization’s AI strategy has an expiration date

I’m starting today with the timeline itself.

And my biggest takeaway from putting it together is this:

👉 “We looked at ChatGPT last year” is not an AI strategy.

If your organization evaluated it 12 months ago, you evaluated a substantially narrower set of business capabilities.

The central question is still:

What can AI actually do for my business?

But we have to keep asking it.

Because the answer does not stay put.

[Momentum AI @MomentumAIonX]

@tgiovanetti I keep saying that... The new OpenAI datacenter in Pike County, OH is 35,000 construction jobs. In New York, apparently, we don't want to put 35,000 people to work.

[Momentum AI @MomentumAIonX]

@SullyOmarr The real joy is doing something like that without it costing you a penny on someone else's software.

[Momentum AI @MomentumAIonX]

@joshelman I miss open seating as a business traveler. For family travel, the new system is a godsend.

And, yes... Though I will say my agent was about four minutes late. I would've been C76 the old way. 🤣

[Momentum AI @MomentumAIonX]

@GergelyOrosz You're not the client any of us are working with. It's easy to surmise here on X that people know what they're doing, but largely they don't. They're busy running their businesses. A well-thought out agentic solution changes their world.