You cannot fool my agents with fake images or manipulated files! 🙅♀️🙅🙅♂️
Introducing scam detection for Codex, powered by @ScamAI_Official x @Aident_AI . Turns result into a review-ready brief with flagged files. Learn more @ https://t.co/Yf3FCBm8RV
20M+ companies. 30K+ subsectors. 70+ data points.
VC research still ends in spreadsheets.
1 thesis. @akta_pro supplies data. @Aident_AI returns 5 sourced companies with risks and next questions.
The outcome: a defensible shortlist. ↓
20M+ companies. 30K+ subsectors. 70+ data points.
VC research still ends in spreadsheets.
1 thesis. @akta_pro supplies data. Aident Loadout returns 5 sourced companies with risks and next questions.
The outcome: a defensible shortlist. ↓
Cheat code for building products:
1. take a product or an open source project that's popular among power users, do one of 2 things below:
- cut 90% of the functionalities, improve on the remaining 10%
- take care of 90% of the set up, so users do not need to mess with set up
2. sell to non-power users, which is 99% of the world.
Now that Muse is finally available in Canada 🇨🇦 , initial thoughts:
- it’s basically OpenClaw and Hermes but without 90% of the setup
- its desktop and mobile apps are slick, best in class (Meta knows how to build messaging apps well)
- they published a blog on how they really focused on security before launch like filtering out sensitive OTP/recovery emails: this is heavily underrated and was what stopped me from using other CoS assistants
Well done @alexandr_wang@natfriedman and team!
If non-technical users love Product A, but technical people dismiss it because “Product B, built for power users, can already do all of that,” Product A may be more likely to win than they think.
You get iPhone Duo. Your agent gets Aident Loadout—and finally, some life skills.
Your agent has a PhD and no hands. Aident Loadout fixes the hands part.
#Aident#Loadout#iphoneDuo#agent
When a tool gets really powerful for a lot of things, people usually have white board problem, i.e., what should i use this tool for? We build skills in Aident Loadout to solve this, no idea how to use the thousands of tools that Loadout support? just head over to https://t.co/cMZOfAhsF8 to pick something and assign to your agent.
Your agent doesn’t need another blank prompt. It needs a head start.
👉https://t.co/k0HsU0XZKx
Aident gives it 400+ expert-built skills for research, outreach, content, ops, and more—proven workflows ready to run from day one.
Pick a skill. Put your agent to work. ⚡
#AIAgents #Aident #Loadout #skills
This is a feature that i really like, i set a cap for all tool calls, when using Aident Loadout in any agent harness, agent will ask for confirmation if the cost of a tool call exceeds my cap. Finally get my AI budget under control.
AI agents are useful. Black boxes aren’t.
Loadout Audit shows every action your agents take — where it came from, whether it worked, how long it took, and what it cost.
👉https://t.co/k0HsU0XZKx
You can even cap the credits used per call!
No guesswork. Just a clear record of what actually happened. 👀
#AIAgents #Aident #AIOps #Loadout
great interview!
the contrarian bet resonates so much with me, for example, i often assign hard tasks to agents before i go to sleep, for those tasks, i'm happy to accept low tps for lower price.
the only thing is that Sail doesn't seem to offer SOTA open source models
Neil Movva (@neilmovva) started his career at Nvidia, working on GPUs and kernels, and has an unusually deep understanding of inference, from software to chips to power.
We spend a lot of time on each of those layers, how they connect, and where the important tradeoffs are.
What makes this conversation special is how detailed it is (like a 401-level class), yet Neil makes it remarkably clear and easy to follow.
Today he runs Sail Research, a company building infrastructure for agents to make tokens as cheap as possible.
We discuss:
- Latency versus throughput
- Why there are no bad chips, only bad pricing
- The end of kernel engineering
- Buying chips and power no one else wants
- New chip architectures
- Nvidia lore + his contrarian view of the company
- Open source and the frontier labs
I learned a ton. Enjoy!
TIMESTAMPS
0:00 Intro
0:38 Building a “Token Factory”
4:21 The Future of Background Agents
13:09 Nvidia and the GPU Stack
23:27 Chips, Memory, and Transformers
36:14 The Future of AI Training Data
44:32 Chip Scarcity and Compute Arbitrage
52:44 Reinventing the AI Data Center
59:01 Power and the “Scavenger Strategy”
1:10:10 Open vs. Closed AI
Claude, Codex, Cursor, Openclaw, Hermesagent can already help fetch data, update databases, and send emails.
But before that happens, you still spend hours dealing with app logins, API keys, MCP setup, OAuth, and scattered credentials.
That’s why we built Aident Loadout.
It’s a tool and permission layer for your agents: connect tools once, keep credentials safe, and let your agents use them across different environments.