- Horizon AI
- Posts
- Use This One-Word ChatGPT Trigger to Expose Flaws in Your Thinking 💡
Use This One-Word ChatGPT Trigger to Expose Flaws in Your Thinking 💡
Anthropic and OpenAI Release Rival AI Models on the Same Day 🔥
Gina Acosta
February 06, 2026
In partnership with
Welcome to another edition of Horizon AI,
Most AI assistants tend to agree with whatever you say. But the real power comes when they challenge your thinking. In today's issue, we share how to create a simple keyword trigger that switches ChatGPT into critique mode to spot flaws in your reasoning
Let’s jump into it!
Read Time: 4.5 min
Here's what's new today in the Horizon AI
Anthropic Releases Opus 4.6 With New “Agent Teams”
OpenAI Launches GPT-5.3-Codex Just 27 Minutes After Anthropic's Opus 4.6
AI Tutorial: Use this one-word ChatGPT trigger to stress-test your thinking
AI Tools to check out
AI Findings/Resources
The Latest in AI and Tech 💡
AI News
ANTHROPIC
Anthropic Releases Opus 4.6 With New “Agent Teams”
Anthropic has rolled out Opus 4.6, upgrading its flagship Claude model with collaborative AI agents, expanded memory, and deeper productivity integrations.
Details:
The standout new feature is "agent teams," which lets multiple AI agents divide a large task into coordinated segments instead of working sequentially.
Anthropic says this parallel structure allows agents to own specific responsibilities while communicating with each other to complete complex workflows faster, comparing it to having a talented team of humans working for you.
Agent teams are currently available as a research preview for API users and subscribers.
On benchmarks, it scores highest on the agentic coding evaluation Terminal-Bench 2.0 and leads all frontier models on Humanity's Last Exam, a complex multidisciplinary reasoning test.
Opus 4.6 also comes with an expanded context window of 1 million tokens and integrates Claude directly into PowerPoint as an accessible side panel.
This new version not only improves on its predecessor's coding skills but also can apply its improved abilities to a range of everyday work tasks as the company aims to reposition it from a model that was highly capable in one particular domain (software development) into one that could be really useful for a broader set of knowledge workers.
TOGETHER WITH VIDMUSE
The First AI Video Tool That Puts the Song in Charge
You don’t need more clips. You need a structure that matches the music. VidMuse listens to your song, then builds a music video around it, so every scene has a reason to be there, and every cut lands with the sound.
It’s your personal AI music video agent, built to turn a track into a real workflow without the usual headache. Use your own music or a team track for demos.
Upload the song; VidMuse builds a blueprint.
It sets timing from the music’s highs.
Pick a style: cinematic, anime, Y2K today.
Create music videos simply by chatting with VidMuse, then adjust easily.
From shot lists to transitions, VidMuse handles the heavy lifting. You focus on the look, the story, and the feeling. It’s the fastest way to get from “I have a track” to “I have a video.”
Get started now 👉 Register now for 1,000 free video credits.
OPENAI
OpenAI Launches GPT-5.3-Codex Just 27 Minutes After Anthropic's Opus 4.6
OpenAI has released GPT-5.3-Codex, its latest coding model that the company says transforms Codex from an agent that can merely "write and review code" to one that can do "nearly anything developers and professionals do on a computer, expanding who can build software and how work gets done."
Details:
OpenAI says that GPT-5.3-Codex is 25% faster than its previous model (GPT-5.2) and that it was the company's first model that "was instrumental in creating itself."
The new model shows substantial gains across multiple industry benchmarks. Most notably, it beats the just-released Opus 4.6 by 12 percentage points on Terminal-Bench 2.0, a significant gap by current AI standards, all while using fewer tokens than any prior model.
GPT-5.3-Codex is available to paying ChatGPT users in the Codex app, CLI, IDE extension, and on the web.
The back-to-back releases added to an already heated week between the two AI giants and also highlight the intensifying competition to capture the enterprise software development market.
AI Tutorial
Use this one-word ChatGPT trigger to stress-test your thinking
Go to ChatGPT and log in.
Click on your account icon in the bottom left corner and select Settings.
Click on Personalization, go to Custom Instructions, and add:
Whenever I type the word '[trigger word]' followed by an idea or argument, I want you to disregard your 'helpful' persona. Instead, act as a Hostile Critic. Your sole job is to identify the 'holes' in my logic. Point out three specific ways my argument could collapse, two assumptions I'm making without evidence, and one counter-argument I haven't considered. Do not worry about being polite, just precise.Now, whenever you have a plan, argument, or decision, type your trigger word plus your idea to run a quick but useful stress test, instead of dealing with the usual 'yes-man.'
You can also simply paste the prompt in a new chat and go from there.
AI Tools to check out
🎮 Playmix: AI game maker.
🎨 GLM-Image: AI-powered professional image editing platform.
📌 Noteihub: Turn GPT chats into to-dos, journals, and reminders.
📹 Crevid: All-in-one AI video & image generator.
💥 Articos: Turn your ideas, landing pages, and messages into structured audience conversations so you can make decisions with clarity, not assumptions
TOGETHER WITH KEEPCART
With KeepCart or Without It — See the Margin Difference
Every checkout leak costs you. KeepCart shows exactly how coupon extensions quietly shrink your revenue — and how blocking them can recover hundreds per order. Your profit margin deserves a before-and-after moment.
Top brands like Bucketlisters and Quince have already seen the difference.
AI Findings/Resources
📉 Why a new AI tool hammered some software stocks this week
📷 Your phone edits all your photos with AI - is it changing your view of reality?
💡 AI is helping funders evaluate more ideas more fairly
The latest in AI and Tech
The company is testing a standalone version of its Vibes app, a short-form AI video platform that until now lived inside the Meta AI app.
Vibes lets users generate, remix, and share AI-created videos in a dedicated feed, similar in format to Reels or TikTok, but focused entirely on AI content.
OpenAI has introduced Frontier, an end-to-end platform aimed at helping enterprises build, deploy, and oversee AI agents as part of their core infrastructure.
The system is designed as an open platform, meaning companies can manage agents created both inside and outside OpenAI’s ecosystem.
The tech giant says it has used its AI tools to help sequence the genomes of 13 endangered species, supporting global efforts to preserve biodiversity.
Working with the Vertebrate Genomes Project and the Earth BioGenome Project, the company is providing funding, infrastructure, and specialized models to accelerate genome sequencing, a process that used to take years and billions of dollars.
The company is rolling out a new Council Mode for Perplexity Max users on the web that routes queries to a group of frontier reasoning models working in parallel, then combines their outputs into a single synthesized response.
Tinder is testing an AI feature called Chemistry aimed at reducing “swipe fatigue,” the burnout many users feel after endlessly scrolling through profiles.
The tool uses AI-driven questions to learn about a user’s preferences and, with permission, can analyze photos in their Camera Roll to better understand interests and personality.
That’s a wrap!
We'd love to hear your thoughts on today's email!Your feedback helps us improve our content |
Not subscribed yet? Sign up here and send it to a colleague or friend!
See you in our next edition!
Gina 👩🏻💻