Two major models in three months: Claude Fable 5 in June, GPT-6 Astra in September. I talk less about what they do today than about what they hint at.
Announcements emphasize power (ah, marketing!). What seems more important to me are two more discreet capabilities: these models retain context over millions of words, and they can sustain a long task without losing the thread. In other words, they’re starting to have solid memory and more endurance. For those who produce brand content, like me, that’s not a detail.
Three perspectives seem to be opening up to me.
Consistency over time.
A brand consists of hundreds of productions that must look alike over the years, with teams changing. This has been our historical weakness. A system that preserves the platform, the guidelines, what has been produced and what has been refused, and why, comes in where we have always struggled.
Exploration becomes affordable again.
Many ideas don’t die in meetings; they die at budgeting. If trying costs little, we can again afford to search before choosing. And looking longer is often what separates a good idea from the right one.
Adaptation stops being a degradation.
Formats, languages, markets: declining it almost always meant impoverishing the original. This inevitability no longer has a place.
That’s the essential: the more capable these systems become, the more value concentrates in the few moments when someone decides: what we’re really looking for, what we keep, what we’re willing to sign. Those moments aren’t delegated, and I don’t yet see what could delegate them. It’s in this direction that we are working at Nextage.ai.
I don’t believe we are heading toward artificial creativity. Rather toward human creativity less hindered. The outlook frankly makes me optimistic for the coming months.
#AI #CreativeAI #Advertising #MarketingInnovation #Nextageai