NXT (@NXTunseen) header
NXT (@NXTunseen) profile picture

Agent of Good Evolving Human; Pro Humanity, aiming for Centre. student of Life . crypto enthusiast since 2017. Jiujitsu. 🧬 activated. protect the children

Sydney, New South Wales
Joined September 2017
Photo by NXT (@NXTunseen)Photo by NXT (@NXTunseen)Photo by NXT (@NXTunseen)Photo by NXT (@NXTunseen)Photo by NXT (@NXTunseen)Photo by NXT (@NXTunseen)Photo by NXT (@NXTunseen)Photo by NXT (@NXTunseen)
ai agents are going to be the next big thing
6
18
1,020
🧠🎮 UNBELIEVABLE! A paralyzed man is playing GTA 6 using NOTHING but his THOUGHTS! Neuralink's Patient 23, known as LeBrain James, is controlling boats, planes and cars through his brain implant. Just imagine what this means for someone who has lost the ability to move freely. The freedom to explore an entire virtual world using only your mind! Elon and his Neuralink team deserve enormous credit for turning something once considered science fiction into reality. This is the kind of technological progress that genuinely restores hope in humanity. Godspeed, Elon! ❤️ Writer: Naz
53
149
11
878
59,900
Sean Strickland calls out young men spending all day playing GTA 6 💀 “I don’t understand you guys. Y’all need to shut off the consoles, go outside and get some p*ssy. This sh*t is embarrassing.”
764
332
96
3,854
461,379
Introducing AIR-O, a model that decodes what a person is attending to directly from brain scans. Using only noninvasive brain data, we can generate an image from scratch in real time that depicts where a person is focusing their attention. This is the first of a new generation of algorithms that can read the shape of thoughts directly from brain activity.
268
573
329
4,596
480,962
Modern life can feel like being in a sci-fi story—all that is missing is the aliens. They probably won’t appear in 2027. Here are ten other things to watch out for bit.ly/3UCxXw9 Illustration: Cleon Peterson
3,238
16,432
7,613
66,491
14,798,407
این مقاله واقعاً دیوانه‌کننده‌ست! پژوهشگرهای چینی مقاله‌ای منتشر کردن با عنوانی که باید تمام سازندگان ایجنت‌های هوش مصنوعی رو به فکر فرو ببره: «پایان Harness Engineering» حرف اصلی مقاله در یک جمله‌ست: شاید دوران طراحی و کدنویسی دستی تمام اجزای اطراف مدل‌های هوش مصنوعی داره به پایان می‌رسه. بذار ساده‌تر بگم: تا امروز ما برای ساخت هر ایجنت، ابزارها، دستورالعمل‌ها، گردش کار و منطق اجرایی رو طراحی می‌کردیم؛ بعد هم همون ساختار رو ماه‌ها استفاده می‌کردیم. اما این مقاله یه ایده متفاوت داره: چرا اجازه ندیم خود سیستم، زیرساختش رو بهتر کنه؟ روش کارش جالبه: ۱. یک مدل کوچیک ۴ میلیارد پارامتری، شکست‌ها و خطاهای سیستم رو بررسی می‌کنه. ۲. برای بهتر شدن عملکرد، ۸ تغییر مختلف پیشنهاد می‌ده. ۳. هر تغییر با مجموعه‌ای از سؤالات و تست‌ها ارزیابی می‌شه. ۴. فقط تغییری باقی می‌مونه که واقعاً نتیجه بهتری گرفته باشه. نکته مهم اینه که مدل اصلی ثابت می‌مونه؛ چیزی که مدام تغییر می‌کنه، هارنس و ساختار اجرایی ایجنته. این ایده به یک مدل یا فریم‌ورک خاص محدود نیست و می‌تونه نگاه ما به ساخت ایجنت‌ها رو تغییر بده. اما بخش مهم‌تر ماجرا: نقش انسان از کدنویسی و دستکاری مداوم سیستم، به طراحی آزمون‌ها و معیارهای موفقیت منتقل می‌شه. یعنی باید دقیق بدونیم از ایجنت چی می‌خوایم و چطور بفهمیم واقعاً بهتر شده یا نه. شاید در آینده مزیت رقابتی ما این نباشه که چقدر خوب برای ایجنت‌ها کد می‌نویسیم؛ بلکه این باشه که چقدر خوب می‌تونیم عملکردشون رو بسنجیم و بهبود بدیم. خلاصه ایده مقاله در یک جمله: قبل از اینکه کد رو بنویسی، امتحانش رو طراحی کن!
141
525
48
3,939
257,247
Prior to #UFC334, Ciryl Gane hadn't won in undisputed UFC title fights 😳 Bon Gamin finally gets it done on the biggest stage against Josh Hokit 👏
84
102
15
1,483
63,458
FACT: 80-YEAR-OLD INVESTING LEGEND PETER BRANDT SAYS #BITCOIN IS GOING TO $600,000 "THIS CYCLE" IT'S COMING 🚀
133
213
12
1,981
58,179
The culprit behind the Microsoft outage has finally been identified…
6
10
2
45
2,297
45.5 万 Star 的逆向怪物 REA 彻底杀疯了! 以前还只帮 AI 扒软件,v6.9.0 直接拉起独立桌面给 AI 录操作、跑验收,连微信小程序和硬件固件都一锅硬拆! 翻完官方更新,这波升级极其凶残: · 桌面任务流水线化:一条命令备好 Linux 构建环境,装入带校验的应用、设好起止节点,一次录制拆出多条桌面任务,自动校验和屏幕裁判直接验收 · 万物皆可硬拆扒光:.NET 单文件、Bun / Deno、Unity / Godot、微信小程序、红白机卡带,一路挖到 ESP 固件、Electron 缓存和离线 NTFS 日志,拿到手直接拆 · 救回爆掉的 Token:以前挂满逆向工具进 Claude Code 上下文直接顶炸,现在按需挑工具、精简结构、分页吐结果,没扫全老实标未知绝不瞎编 · 主流工具直接开干:Claude Code、Codex、Cursor、LM Studio 原生接管 搞逆向、桌面 Agent 和折腾 Claude Code 的兄弟建议直接马住! 更新地址:github.com/morluto/rea/relea… #开源 #逆向
65
149
4
1,107
73,965
NXT retweeted
UNCENSORED DeepSeek, GLM and Kimi in ONE API 💀 found cheap some abliterated open models with zero data retention models you get: - GLM 5.3 Flash Abliterated - DeepSeek V4.1 Flash 1017 Abliterated - Kimi K3.1 Pro Abliterated (saw it in the roadmap) >works in Claude Code and Cursor top up and pay per token, no monthly fee grab your API key at lutheos.com base url lutheos.com/v1 No idea how long these models stays up
9
5
1
96
32,207
DeepSeek V4.1 Flash 1017 and HY 4 + more models for free you can get 20M tokens daily - GLM 5.3 Flash - DeepSeek V4.1 Flash 1017 - HY 4 - GPT-6 Luna - + more setup is very simple: just sign up, get your API key, go to the playground, select a model, and start chatting. you can use it also with coding tools and APIs like Hermes, Cursor, Codex, Claude Code, OpenCode, and more. Try it here: apinex.bond
40
23
2
349
31,644
NXT retweeted
wait i just found UNCENSORED DeepSeek and GLM 😳 one API with abliterated open models and zero data retention - GLM 5.3 Flash Abliterated - DeepSeek V4.1 Flash 1017 Abliterated - DeepSeek V4.1 Flash 1017 - GLM 5.3 Flash how to use it: 1. go to lutheos.com and sign in 2. create an API key 3. top up 4. pick an abliterated model > works in Claude Code and Cursor base url lutheos.com/v1
9
6
117
8,536
NXT retweeted
ai engineers thank me later > insideout.sh
80
236
39
2,614
405,047
How to build a ChatGPT trading bot for stocks, no TradingView needed: Step 1: Get the tools Open ChatGPT desktop and switch to Codex. This is the agent. Make a free account at tradingkit.com. Go to “My Account”, copy the line of code, paste it into Codex and say “install this globally”. It does the backtesting. Open a paper trading account at alpaca.markets. Free, and no real money until you’re ready. Step 2: Check it works Prompt “Backtest a simple 20/50 moving average cross on SPY daily, just to confirm we’re connected” Step 3: Find the strategies Prompt “You are an elite quant researcher with access to the tradingkit tools. For the next 2 hours, search Google Scholar, Quantpedia, TradingView community scripts and YouTube for stock and ETF strategies. Build and backtest each one on SPY, QQQ, AAPL and NVDA on the daily and 4 hour. Fees and slippage on. Trade only at the open of the next candle. Optimise on two thirds of the history and validate on the final third. Only keep strategies with over 100 trades, max drawdown under 20%, profit factor above 1.1 and that beat buy and hold. Save every one that passes to a file called survivors with the full rules” Mine tested 1,368 strategies in 2 hours. 277 passed. Step 4: Don’t trade any of them yet Click Incubate in tradingkit and leave the best ones for 1 to 3 months on data they’ve never seen. 99% fail here. I’ve tested over 2,000 strategies in 5 years and I run 7. Step 5: Watch the gap Stocks close at 4pm and reopen with gaps a bot can’t plan for. Keep leverage at 1x and position size small until a strategy has lived through a few earnings days. Step 6: Go live Create Alpaca API keys with trading permission only. Connect them through trigger.trade, which encrypts the keys on your machine. Ask Codex: “Set up an alert for the survivor strategy through tradingkit and send entry, stop loss and take profit to this trigger.trade webhook in one signal” Step 7: Diversify One bot per ticker. Mix trend following with mean reversion. Losers get covered by winners and the equity curve flattens out. Then leave it alone.
10
1
2
26
2,196
NXT retweeted
Ahora tengo una necesidad muy grande.
80
792
59
9,323
556,112
JUST IN: FOMO FOUNDER SE YONG SAYS THEY DO NOT WORK WITH CREATORS WHO RUG THEIR FOLLOWERS “THE CREATORS WE WORK WITH ARE FOCUSED ON EDUCATING USERS, NOT RUGGING THEM”
224
63
49
720
184,990
One of crypto's largest KOLs has been secretly extracting millions from his audience We’ll expose him soon Unfortunately, he’s just a symptom of a much bigger problem Social trading has brought out the worst in KOLs. It was designed to keep them honest by making every trade public. In reality, the opposite happened In just a few weeks, we exposed a number of "top social traders" using the same tricks: • Public wallets to attract copy traders • Side wallets to dump • Callouts to farm rewards • Bundled tokens to hide supply And there are plenty more doing the exact same thing Why? Because our industry incentivizes it 1) Launchpads allow it Most launchpads today are obsessed with one thing: volume. Millions in callout rewards, affiliate links, and top leaderboards. Almost nothing on detecting bundles, side wallets, or shady traders. Scams drive volume. Launchpads have no reason to stop it (other than actually caring about their users) 2) Social trading incentivizes it Side wallets have become a business model. Before social trading, only big KOLs could do it. Now, anyone who lands on a leaderboard can monetize by dumping on their copy-traders. The temptation is real 3) CT normalizes bundle culture “Everyone bundles. It's how it's done" - an excuse we hear every time we expose a KOL. Supply control isn't the problem. Hiding it is. If it's so bullish, why hide it? Spoiler: because you're planning to sell *** Launchpads are paying millions to dishonest KOLs. It's time they start taking it seriously Audit the top traders. Expose their side wallets. Ban repeat offenders We have the tools. What's missing is the willingness to use them No one's coming to save us
1,702
919
693
7,389
1,291,707
I'M SHUTTING DOWN THE INTERNET!!!😠🤣🤣🤣
🤖 Made with AI
1,358
4,493
489
25,244
1,211,755
Focus on your goal, everything else is just distraction
67
761
69
9,683
711,346