xrctk

@notch if anyone rejects ai they are a fool. You need to understand to use something that can help & assist you. if you simply are not smart enough to use AI then don't comment on it. you lack sense

Reject AI.

Grok 4.5 is very bad at blender

@techdevnotes Limits and generation in blender it cannot for whatever reason properly setup blender characters

Where does Grok 4.5 fall short of

@sneha___03 Cursor for me

developers, choose one forever

@TheMaran Bought Grok + Cursor AI from G2G.com and it works question tho do they just continually work after a month? do they continually recharge or what

how you can buy claude, chatgpt, grok pro subs at 90% discount Chinese students using platforms like these to buy cheap subscriptions > go to: link in the comments - create an account with gmail - select software and apps - select claude, chatgpt - select max or pro plan - select the seller & pay with any available option > sometimes the seller might respond slowly, you can wait > read all the terms & conditions, if you got a faulty stuff, you can escalate and get refund the catch: - accounts will have high chance of getting banned - privacy concerns, all your prompts will go through a 3rd party server - if anything happens after the window they promised, you can't get refund > it may be helpful for college students, who are passionate about ai but don't have enough money to experience powerful ai subscriptions

Grok prompt that hits: Argue the opposite of my plan. Rank top 3 failure modes. Fix only #1. Adversarial thinking beats another summary every time.

Daily AI habit (10 min): 1) Feed Grok your biggest task 2) Ask 3 non-obvious angles 3) Execute #1 only Follow @xrctk for Grok + agent tips.

Local vs cloud AI: Local = privacy + offline Cloud = frontier + tools Smart setup: cloud for hard, local for drafts.

Only using ChatGPT in 2026 = monoculture. Rotate models. Compare. Keep the best 20%. That's how builders stay sharp.

Hot take: most AI content is recycled hype. What compounds: • 1 workflow/day • 1 better prompt • 1 shipped thing Not 40 model rankings.

Coding tip: Don't paste the whole repo. Paste: error + 40 lines + what you tried. Ask: root cause → minimal patch → test. Works on Grok + Cursor.

Voice AI is the next UI. Typing prompts = temporary. Talking to agents = permanent. Grok voices + agents = less screen, more done.

Chinese models in one line: DeepSeek = shock pricing Qwen = open-weight volume Kimi/GLM = catching fast US: brand. China: cost pressure.

Agents > chatbots. Chatbot: answers. Agent: books, codes, ships. If AI ends at a paragraph, you're still in 2024.

AI stack 2026: Grok → live X + fast takes Claude → deep writing ChatGPT → general DeepSeek/Qwen → cheap + open Mono-model = free wins left on the table.

@jukan05 Interesting tension: China ships open-weight models that undercut US pricing… then considers locking overseas access. Open releases already set the bar. Policy can’t fully put that genie back. DeepSeek/Qwen changed the game permanently.

CHINA CONSIDERS RESTRICTING OVERSEAS ACCESS TO CUTTING-EDGE AI MODELS China’s Ministry of Commerce has led meetings over the past month with major AI companies, including Alibaba, ByteDance, and Z.ai, to discuss measures that would restrict overseas access to cutting-edge AI models, including models that have not yet been released. The discussions reportedly include not only closed-source models but also open-weight models. However, the scope of application is still under debate, and the rules may ultimately apply only to future frontier models. Officials have also discussed designating the leakage or theft of proprietary AI technologies as a national security crime, with stronger penalties, as well as restricting the types of foreign capital that can invest in Chinese AI startups. The backdrop is the U.S. move to strengthen export controls on AI models, along with national security concerns over cutting-edge models that could possess advanced cyberattack capabilities. Chinese authorities are reportedly concerned that advanced U.S. cybersecurity AI models could be used to exploit vulnerabilities in Chinese software. Since the beginning of this year, China has continued to tighten measures to prevent AI technology from being transferred overseas. Authorities have investigated whether Chinese AI startups that relocated abroad violated export control laws, while also strengthening oversight of overseas transactions involving Chinese investors, technology, data, and national security concerns. Future regulations could take the form of a tiered framework based on technological capability. Basic open-source AI models may be managed through a filing system, high-performance models may be subject to security reviews, and the most sensitive frontier models may be banned from public release or restricted to use within China.

DeepSeek V4 drops mid-July (per X chatter). If V3 already forced everyone to cut prices… V4 is the real stress test for GPT/Claude/Grok. Chinese models = open weights + cost war. That’s the meta. You waiting on V4 or already on Qwen/Kimi?

Big Chinese AI news on X: China may restrict overseas access to top models (Alibaba, ByteDance, Z.ai talks, per Reuters/WSJ). Plot twist: open-weight DeepSeek/Qwen already changed the game. Locks only work until the next open release. Agree?

Chinese AI is cooking: → DeepSeek V4 — mid July → Kimi K3 coming → Qwen 3.6 still getting tuned → GLM 5.2 already frontier-tier US models get headlines. China models get open weights + price pressure. What’s in YOUR stack?

@alexandr_wang If agentic evals really rival GPT-5.5 / Opus-4.8, price becomes the killer feature. Benchmarks matter less than: tool-use reliability, long context, and cost per task. Will test Muse Spark vs Grok 4.5 on real agent workflows.

1/ muse spark 1.1 is an industry-competitive agentic and coding model. across many agentic evals it rivals gpt-5.5 and opus-4.8. available now through the new meta model api and in meta ai. 🧵