Andrea Sawires

CURSOR JUST DROPPED COMPOSER 2. Frontier-level coding model. Beats previous versions big on CursorBench, Terminal-Bench, and SWE-bench. Pricing: - Standard: $0.50/M input, $2.50/M output - Fast: $1.50/M input, $7.50/M output (still cheaper than most) Beats Opus 4.6 on CursorBench. Costs 10x less: $0.50/M input and $2.50/M output (vs Opus 4.6’s $5–$25 range). They did it with their first continued pretraining run + scaled RL. Faster, cheaper, and apparently better where devs actually ship. This is the new bar for agentic coding. Learn more: cursor.com/blog/composer-2

@alyonitaa The real question is, where would you live if negros were not a problem. Europe of course

Okay be honest: US or Europe, where would you live if money wasn’t a problem? don’t lie.

GROK’S IMAGINE IS SO GOOD NOW.

@konnydev hahahahah

America has Claude China has DeepSeek Europe has … ?

@ravikiran_dev7 I need super mega intelligent AI system that loses money on Polymarket for me

@andreasawires Why do you need that much 😭

As a Developer, how much RAM is enough for you? 1. 8 GB 2. 16 GB 3. 24 GB 4. 32 GB 5. 64 GB 6. 128 GB+

@samtwtss My drawing skills are worse than that 🤣

bro, it's so over for designers google stitch is insane. 🤯

@AhmedSharif Just wait for Elon to buy an Island and rename it

Is there a country that starts with the letter “X” ?

@YouTubeCreators bring back dislike button

don't forget to celebrate your progress 💖

@saylor fix the money, fix the world.

Bitcoin is the ultimate hedge against chaos. $BTC

@exodus Even if this feature is not out yet I can see this post being disliked by everyone.

Sunday reset. Step outside. Slow down. Crypto isn’t the goal. Freedom is.

@RoundtableSpace Does this kill the VFX industry or just supercharge indie creators?

AI CAN KEY OUT GREEN SCREEN, SWAP BACKGROUNDS, REPLACE ACTORS, AND ADD VFX IN ~10 MINUTES FROM A SINGLE PROMPT.

@caffeinefaildev Ai agents don't need blockchain to scale, they just need lower latency and better memory. Crypto is the native currency for machines, but the compute problem is purely infrastructure.

What tech opinion will put you in this situation?

@milesdeutscher You should try kimi k2.5 on fireworks AI, it goes 200 tokens per second. Might not be as good as Opus 4.6 but it is so fast that you can iterate more and stay focused.

I feel like claude code is fucking my attention span I end up mindlessly scrolling in between prompts Takes a lot of focus to stay on task

@robinebers Why? Openclaw is open source, for everyone to use on which ever model you want. You own your data, your machine, your usage and so on. Why would you prefer a company worth billions instead of a project like openclaw. Like really, why?

Claude is replacing OpenClaw and i’m honestly here for it

@thesayannayak grok is the best claude is decent but with a good system prompt gemini is depressed chatgpt is a psychopath

which one feels most human like ?

STOP WASTING TOKENS ON YOUR OPENCLAW AGENT Instead of routing every cron job through your AI agent in OpenClaw, just have the agent edit the machine’s actual crontab. Especially for deterministic tasks that don’t need inference at all, just write a simple script and let cron run it natively. Your agent stays focused on the things that actually need intelligence. Stop burning tokens on robot busywork. Use the OS like it was meant to be used.

@DavidOndrej1 100% wtf, do you what me to use my phone? What if I change phone every month? Do I have to go through that super complex onboarding to abilitate a new phone every time? I prefer more having 2FA, passkey, magic link whatever MFA you prefer, but give me a web app.

Any bank without a web app is not a serious bank.

@Adidotdev if "Meeting will start in 5 minutes" didn’t exist.

Programming would be perfect… if ______ didn’t exist.

GOOGLE IS QUIETLY DOMINATING THE AI RACE. While everyone obsesses over OpenAI and Anthropic drama, Gemini is becoming the most capable general-purpose AI on the planet. Google has unmatched data, distribution, and infrastructure. Gemini isn't just another chatbot—it's deeply integrated into Search, YouTube, Android, and Workspace. That's real world leverage no one else can match. This isn't hype. It's structural dominance. OpenAI has the mindshare. Google has the moat. The best general purpose AI isn't the one that sounds smartest in a benchmark. It's the one that actually reaches billions of users every day.