who's tested it already? Claude Opus 4.7
in the announcement focus is on working reliability:
> model should handle long tasks more confidently
> follow instructions more accurately
> and double-check its result before responding more often
based on the release card the upgrade over Opus 4.6 looks noticeable as we see
Anthropic positions 4.7 as a model for long autonomous work where you need to keep task flow, check yourself along the way and need less manual oversight
feels like the jump from 4.5 to 4.6 objectively better at keeping context, actually breaks instructions less often, especially the “don't do this” parts
basically no revolution but pre-nerf Opus 4.6 was more than enough for everything i need now and if this is the same but with known bugs fixed i'm into it...
ANTHROPIC RELEASED CLAUDE OPUS 4.7
it's actually a beast, but not everywhere
some things aren't as cool, eats through limits real fast
rn you can use it everywhere
but i'm waiting for claude sonnet 4.7
tool calling and chain-of-thought were improved
unleash the agents
CT account and a dream
in a month my account will be a year old and here’s what I realized
I came here because of a @MrUsertyt article, he doesn’t run Twitter anymore as far as I can see, but back then he published an article saying Twitter is cool and everyone needs an account here
I came during the infofury period, but now I’m just a content creator and happy to meet new people, I also met @MarvinHarryP brother here haha
thanks everyone, I’m glad to be here and I’ll be here for a long time, and soon I’ll meet even more people
why did you come to Twitter?
why risk liquidation trading OIL when you can farm funding with delta-neutral setups
example: WTI with yields of up to 2000% APY
i built a tool specifically for you that tracks all such opportunities across tradfi futures markets on 50 perpDEXs and CEXs
https://t.co/OrPRFPaKXt
web3 marketers won’t be jobless
cos there’s one of the most crucial tasks ai doesn’t solve yet - it’s the distribution question
now entry barrier for creating your own product is lower than ever
but the question how to get in front of your product’s potential user’s eyes ai doesn’t solve
@Anomico5 fair point
yeah, early stage is risky, but there’s a difference:
Speculation = betting on narrative, hype and project with no utility
Utility = betting on solving a real problem
Dragonfly shifted from the first to the second
hope i got you right
Utility > Speculation
you’ve already seen Dragonfly raised a 4th fund for $650M
2026 will be the shift from speculative projects to real ones
focus of venture capital’s already on:
> stablecoins
> payment infrastructure
> rwa
hope you’ll find your spot in the next cycle