GMP

@ItsAlexhere0 Not even a hot take, this is the truth. Anyone can test this with just one prompt

Hot take: $20 Codex Plus is way more useful than $20 Claude Pro

@mweinbach This is true

I'd rank it ChatGPT > Claude > Gemini Funny how the rankings change depending on use case, basically nobody will own the market overall

@bridgerloftin @MelkeyDev You’re Right I have been building a very complex project for a while now, these days GPT fixes issues in it flawlessly, Claude on the other hands is been doing it own things

@SemGM_ @MelkeyDev Agreed, I've found that when I give Opus 4.7 in claude code a coding task it says its done. When I audit the code with Codex, it finds errors and security flaws. When Codex writes the code, Opus says its perfect.

Opus 4.7 > Codex 5.4? 🤔

@JLarky You clearly didn't watch the videos, just took the thumbnails to chase cloud lol

Absolute cinema

@benln Building a SAAS

Who's building over the holidays? Will add you to a group chat I started on X.

@ZacksJerryRig Use Gemini and make that happen

About a year ago a local software company bid me $100k - $150k to create custom manufacturing software for my wheelchair factory. Fast forward a year - they still aren't finished with the original scope of work - and now want an *additional* $100k because *they* went over budget. I've already paid $150k. What would you do in this situation?

@ZECKCOIN @Zcash 0xeb1e2f56a2a395ab7100d9cbea9f847e9c4eb0d7

@cjzafir I think there are levels and stages to what everyone calls vibecoding, and i can tell you when you know how to use the tools and you act as the project manager and let the AI be the developer, the experience is far more different

vibe coding:

@Rainmaker1973 Check Nano banana fix this for you lol

Artist Milly Bampini is so talented and skilled that she doesn't need Photoshop, only paint and brushes. [📹 millybampainti]

@Starwatcher_vc @cursor_ai @cursor_ai i had the same issue, please you need to review this, since when the pricing changed ?

@cursor_ai No mention of new pricing. Not in email. Not tw. Not in this Changelog. Just a notification after update that I have spent nearly 100$ and reached some rate limit I haven't been aware of while using a yearly Pro plan.

Cursor can now make to-do lists, search PRs, and queue messages. We’ve also made Tab much faster.

@Ryan_Kubanka @bindureddy I wanted to comment the same thing

@bindureddy Agree with everything except laptops. I absolutely love my Mac. They will flop with the glasses as well. Interested to see what’s to come for the latest Siri news

I think it's time to sell $APPL - Vision Pro was the biggest flop ever - For the 1st time, I haven't upgraded my iPhone - Laptops are meh... - Abandoned autonomous cars - No word on Apple robots WHAT EVEN!! 😱😱

@BartWarrot @iannuttall If you have a rules around the dB it always keep context, try that I use both of them and sometimes Google fix a bug that Claude is been struggling with, the opposite happens as well sometimes

@iannuttall Definite disagree! Gemini 2.5 is way better in keeping consistent with database scheme etc, Claude keeps inventing stuff and forgetting names, even when it has access to DB MCP for some reason. But could be different because I work in Dutch mostly?

Hot take 🌶️: sonnet AND opus 4 are both better than new gemini 2.5 06-05

@sundarpichai @lmarena_ai @cursor_ai wen

Our latest Gemini 2.5 Pro update is now in preview. It’s better at coding, reasoning, science + math, shows improved performance across key benchmarks (AIDER Polyglot, GPQA, HLE to name a few), and leads @lmarena_ai with a 24pt Elo score jump since the previous version. We also heard your feedback and made improvements to style and the structure of responses. Try it in AI Studio, Vertex AI, and @Geminiapp. GA coming soon!

@cursor_ai I’m not sure if I’m looking in the wrong place, but I can’t restore the code from before the prompt. This feature used to be so helpful—it allowed us to recover the code when the AI made a mistake.

People have no idea what @RujiraNetwork is poised to become First, @THORChain is enabling smart contracts and thus will become an L1. Very bullish in it's own right But the true 0 -> 1 innovation is that Rujira's DeFi suite will be accessible to every chain that TC supports This means that native $BTC will have access to every DeFi primitive. TC x Rujira has solved BTC DeFi But this is also true of $BCH, $LTC, and even $DOGE. And soon $XRP For devs, this means that deploying one contract can access every chain that TC supports. 🤯 $RUNE $RUJI

@ericzakariasson I have been using this before the update and it been great it will be great though for cursor to have a way to tell the AI which directory is which, because multiple times I had the AI creating the backend directory inside the frontend directory

multi-root workspace support is live in cursor for developers that work across multiple repositories, just add the codebases to the same workspace and cursor will index them all

@lightfi_crypto @cursor_ai Really, this is how mine looks like, I have tried to update without success I am on mac

@SemGM_ @cursor_ai It’s there for major updates I think not minor patches

of all the features that @cursor_ai cooked in this release, this inline changelog might be my favourite

@jpthor @HooriRn @THORChain Where to buy

@THORChain redemption arc begins! The only evolving DeFi project on earth.