<![CDATA[Think4.Me]]>https://www.think4.mehttps://substackcdn.com/image/fetch/$s_!S2o-!,w_256,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F88b649ac-b5b6-4fe4-9da0-bc528b4fd0ed_800x800.pngThink4.Mehttps://www.think4.meSubstackMon, 11 May 2026 00:28:18 GMT<![CDATA[FEB 18: Weekly AI Links and Prompts]]>https://www.think4.me/p/feb-18-weekly-ai-links-and-promptshttps://www.think4.me/p/feb-18-weekly-ai-links-and-promptsWed, 18 Feb 2026 16:42:38 GMT Use Cases
Learnings from 800+ GenAI and ML use cases
https://www.evidentlyai.com/blog/gen-ai-applications

Techniques and Training
13 Foundational Types of AI Models
https://www.turingpost.com/p/13-foundational-types-of-ai-models

Impacts: Business, Tech, Partnerships
OpenClaw Creator Peter Steinberger Joins OpenAI, OpenClaw To Continue As Open-source Project
https://officechai.com/ai/openclaw-creator-peter-steinberger-joins-openai-openclaw-to-continue-as-open-source-project/

DALL-E 3: Scheduled for retirement on February 18, 2026. Users are urged to migrate to gpt-image-1-mini to avoid service disruption.

Models/Agents
Grok 4.2 from xAI (beta)
"4 competing agents; not in Microsoft Foundry yet"

GLM-5 from Z.AI
"GLM-5 is designed for complex systems engineering and long-horizon agentic tasks. Open weights and scores very high."
https://z.ai/blog/glm-5

Azure Copilot from Microsoft
"Get assistance in Azure"
https://azure.microsoft.com/en-us/products/copilot
https://learn.microsoft.com/en-us/azure/copilot/overview

Research/Papers
Prompt Repetition Improves Non-Reasoning LLMs
https://arxiv.org/pdf/2512.14982

Self-Feedback–Driven LLM Reasoning
https://arxiv.org/pdf/2602.09000

APIs/Coding
Building a LLM-from-scratch
https://github.com/rasbt/LLMs-from-scratch

Little Prompts:
...
Where could an AI system fail in this scenario?

]]><![CDATA[Weekly: AI Links and Little Prompts]]>https://www.think4.me/p/coming-soonhttps://www.think4.me/p/coming-soonFri, 13 Feb 2026 18:46:37 GMTAI Hours: Links and Prompts

Use Cases

Transformative AI starts with clear use cases

https://aws.amazon.com/ai/generative-ai/use-cases/

Techniques and Training

Calling AI a Stochastic Parrot is wrong - Geoffey Hilton - known for his work on artificial neural networks, which earned him the title of ‘Godfather of AI’

“Calling it stochastic parrot lacks an understanding of the mechanisms; it is not statistics.”

https://medium.com/@jk1849716/from-statistical-parrots-to-topological-cognition-beyond-feature-spaces-in-generative-models-54b44dd875e3

Impacts: Business, Tech, Partnerships

From magic to malware: How OpenClaw’s agent skills become an attack surface

“One of the top skills for OpenClaw is malware”

https://1password.com/blog/from-magic-to-malware-how-openclaws-agent-skills-become-an-attack-surface

Models/Agents

Anthropic’s Opus 4.6 built a C compiler with team of parallel Claudes

https://www.anthropic.com/engineering/building-c-compiler

Qwen3-Coder-Next

“Alibaba’s open weights model; 80B with 3B active; scores better than previous 235B model”

https://huggingface.co/Qwen/Qwen3-Coder-Next

Research/Papers

Scaling Long-Horizon LLM Agent via Context Folding

“Context Folder and Agent Folding are emerging useful techniques

Context folding is a pattern in AI system design where you compress, transform, or selectively carry forward prior interaction state into a structured representation that fits within the model’s context window — while preserving what matters.”

https://arxiv.org/pdf/2510.11967

Recursive Language Models: the paradigm of 2026

https://www.primeintellect.ai/blog/rlm

APIs/Coding

Palantir’s Foundry and Ontology Architectural Patterns for Data Engineers

“Taking AI & Data to a higher level”

https://medium.com/@sachtekchanda90/im-a-data-engineer-what-is-palantir-foundry-7e7568aed87e

https://medium.com/@sachtekchanda90/ontology-catalog-design-in-palantir-foundry-part-1-21904cebd7d3

Little Prompts:

...

Show the decision points.

Subscribe now

]]>