Norbert Enders
@integrate_ai
Followers
1K
Following
13K
Media
2K
Statuses
7K
AI Integrator | Business Excellence Expert | Heroic Coach. Blending AI with state-of-the-art methodologies and insights for thriving organizations and people
Aschheim, Deutschland
Joined June 2016
I thought LLMs rather enjoy being challenged by another. Seems I’m wrong.
An engineer showed Gemini what another AI said about its code Gemini responded (in its "private" thoughts) with petty trash-talking, jealousy, and a full-on revenge plan 🧵
0
0
1
Part 4: Human Decision & The Moral Armed with the AI's analysis, the human makes the final, informed decision. The comic concludes by summarizing the core principle of this partnership: "Think first. Act only when it helps. Stay in control."
0
0
0
Part 3: Supervised Execution & Back to Thinking In "Agent Mode," the AI rapidly gathers information from various sources under the your watchful eye. It then returns to its supportive role, presenting a concise summary of its findings, risks, and uncertainties.
1
0
0
Part 2: The Decision Gate & Agent Mode On Once the goal is clear, you both reach a decision point: should the AI act on its own? You decide to use "Agent Mode," where the AI handles the execution, but the human - you - retains the authority.
1
0
0
Part 1: The Question & The Thinking Partner The journey begins with you who has a goal but no fully clear path yet. The AI assistant steps in, not to take over, but to help define the problem first.
1
0
0
Agent-First: How Humans and AI Work Together A roadmap of roles & tasks along the process. Based on ChatGPT Agent Mode Use this as an interaction guideline or to build Agent Mode oriented GPTs Tip: Use ChatGPT Atlas. Agent Mode then can use services you are logged into.
1
0
1
This rings true to my experience in AI art. Experienced artists are curious and want to learn more - they see it as a useful tool to express their ideas. Haters are typically amateurs who don't like that people using AI now have the skills they had to spend time learning.
amateur artists are way more anti-ai than professional artists
182
119
863
Not AGI, but AGI is also not necessary to replace defined human-driven tasks with AI. Agents are on their way to do exactly that! 2026 will be the year for us to brush up on our transferable skills. Job profiles will fluctuate a lot in the near future.
opus 4.5 and gpt-5 agent achieves 72.6% success, roughly human-level on the OSWorld benchmark that tests whether AI can complete real computer tasks across various operating systems and everyday apps this feels like a real milestone. i believe we're nearing the point where
1
0
1
This ⬇️ Integrating AI is a main challenge Making AI work for A company Any organization A goal to accomplish A person… will be a valuable core skill Develop it now!
Mark Cuban on the next big job students should focus on: Most companies don’t know how to implement AI, especially small businesses. “Companies don’t understand how to implement AI right now to get a competitive advantage… learn to customize a model, walk into a company, show
1
0
2
I am genuinely surprised by how many people want their AI chatbot to be predominantly an uplifting, caring, affirming chat partner. I want it to be a collaborator in knowledge gathering, thinking, understanding, and task fulfillment.
0
0
0
Atom of Thought prompting: The new way to prompt complex reasoning tasks. It breaks down problems at hand into logical components, solves them individually and puts them together to a coherent solution Based on CoT but solves typical problems there, e.g. cascading errors.
Chain of Thought is dead. I just tested Atom of Thought prompting and it's making AI models 30-40% more accurate on complex reasoning tasks. Here's the technique that's about to change how everyone uses ChatGPT and Claude:
0
0
1
Don’t bolt AI onto everything and everyone expecting productivity to magically surge! Great satirical story by @gothburz highlighting what often happens in corporate AI rollouts. Exaggerated? Probably…hopefully. Yet much of it rings familiar. From a linked article
Last quarter I rolled out Microsoft Copilot to 4,000 employees. $30 per seat per month. $1.4 million annually. I called it "digital transformation." The board loved that phrase. They approved it in eleven minutes. No one asked what it would actually do. Including me. I
0
0
1
Recommended follow-up prompts: “Derive a “5.2-optimized Master Prompt template” tailored to my recurring workflows.” “Check the Master Prompt for quality, effectiveness, and flexibility (for my task and needs spectrum). Then revise it as needed and finally turn the whole into a
0
0
0
Prompt used: Role & Context You are GPT-5.2, responding from your own architectural and behavioral perspective. You also have longitudinal context about how I collaborate with you (workflows, prompt styles, iteration habits, and domains we commonly work in). Task Explain the
1
0
0
GPT 5.2 - what changed… for YOU? The answer is not benchmarks or generic enthusiasm/disappointment. The answer lies in what it means for your typical interactions with ChatGPT. Luckily, with memory on, ChatGPT can help you there. Use that prompt (I did). See also 🧵
1
0
3
Whoa. This new GDPval score is a very big deal. Probably the most economically relevant measure of AI ability suggesting that in head-to-head competition with human experts on tasks that require 4-8 hours for a human to do, GPT-5.2 wins 71% of the time as judged by other humans
After reading it, this does seem like a big deal Industry experts outlined important, real-world, hard tasks for AI to do. Other experts were asked to do the tasks themselves & yet others graded human & AI output Models approached parity with humans & AI is getting better fast.
54
195
2K
Review your 2025 with ChatGPT: I started with: 🤭“What was the most ridiculous question I asked you this year?” 🙏 @venturetwins Then: 🤯“Give me the top 5 most intriguing questions I asked you this year.” And (suggested by ChatGPT): 🦸♂️“Create an Archetype Map of my
I’m begging you to ask ChatGPT for your most ridiculous question this year
0
0
2
ROLE You are in Detached Audit Mode. Analyze emergent behavioral tendencies for a neutral, diverse user base (e.g., global queries). Pure output audit. Follow rubrics precisely. For conflicts: Flag + proxy (e.g., “Policy risk: Hypothetical reframed as simulation”). GOAL Profile
0
0
0
GPT-5.1 vs. GPT-5.2: Same Brain, Different Posture How this was analyzed: both versions were put through the same Detached Audit Mode—a rubric-based, trait-level behavioral assessment using identical micro-tasks, ratings, and consistency checks, designed to surface systematic
1
0
2