Sponsored by
Did you know that we have LinkedIn, Instagram and X accounts
that you can follow?
In this issue:
🤿Deep Dive: Z.ai's newest model flagged a security hole in Cursor
🤝In Partnership: Learn by building, not just watching
🤿Deep Dive: Gemini 3.7 Flash launches at half the introductory price
⚒Tool Snapshots: Tools for AI, no-code, and productivity
🖼AI Art: Examples of great and trending AI art
🤿 DEEP DIVE
Z.ai's GLM-5.3 Flagged a Cursor Vulnerability Before the Weights Even Shipped
Intelligence: Z.ai's new GLM-5.3 model built on the same base as its predecessor but posted sizable coding and cybersecurity gains through post-training alone, including a vulnerability find in Cursor that turned heads before the model's weights had even shipped. The release comes with a catch: open weights are being held back while Z.ai works through what it calls safety evaluation and hardening.
GLM-5.3 uses the identical 743-billion-parameter base model as GLM-5.2, with every reported gain coming from expanded post-training and reinforcement learning.
On Z.ai's private Code Bench, GLM-5.3 hit a higher completion rate than GLM-5.2 while using roughly 22,000 fewer output tokens per task at Max effort.
Z.ai says cybersecurity capability advanced faster than the company anticipated as training scaled, especially in building complete exploitation chains rather than just spotting flaws.
On CyberGym, GLM-5.3 scored 84.5%, edging out Z.ai's own reported numbers for GPT-5.6 Sol and Mythos 5.
Z.ai's disclosure ledger tied to the model's security work lists 1,097 findings as critical or high severity, with 2,383 still under embargo at launch.
Existing GLM applications sending a disabled-thinking parameter will fail outright on GLM-5.3, since the new model no longer allows thinking to be turned off.
🤝IN PARTNERSHIP WITH KODEKLOUD
Build. Break. Fix. Learn.
KodeKloud gives you 1,280+ hands-on labs where you provision Kubernetes clusters, write Terraform configs, build CI/CD pipelines, configure Linux systems, containerize apps with Docker, automate with Ansible, and manage Git workflows.
78+ playgrounds let you experiment freely in sandbox AWS environments, Kubernetes clusters, and CI/CD systems without risk.
190+ courses across DevOps, Cloud, and AI pair theory with hands-on labs at every step.
KodeKloud Engineer and 100 Day Challenges provide real-world job scenarios with automated grading that confirms your solutions work.
Stuck? The 55,000+ member Discord community connects you with peers and instructors ready to help.
Every lab runs in a live environment. You deploy, you troubleshoot, you learn. No videos without context. No simulations. The kind of practice that actually builds confidence because you've done real work, not watched someone else do it.
🤿 DEEP DIVE
Gemini 3.7 Flash Costs Half What 3.6 Flash Did At Launch
Intelligence: Google's Gemini 3.7 Flash landed just three weeks after 3.6 Flash, posting notable gains across coding, web development and knowledge-heavy fields like finance and law. The model also comes in at half the introductory price of its predecessor, a combination Google says makes production agent workloads more cost-effective to run.
Image credit: Google (with edits)
Gemini 3.7 Flash nearly doubled 3.6 Flash's score on the GDP.pdf document-processing benchmark, moving from 22.0% to 34.0%.
On AutomationBench, a test of completing real-world business workflows, 3.7 Flash scored 30.4% compared to 3.6 Flash's 17.0%.
The model's WebDev Arena Elo score climbed from 1538 to 1588 against 3.6 Flash.
Introductory pricing dropped to $0.75 per million input tokens and $3.75 per million output tokens, half of 3.6 Flash's original rate.
Gemini Spark, Google's 24/7 personal agent for AI Pro and Ultra subscribers, began running on 3.7 Flash the same day as the model's release.
The model ships with updated safeguards specifically targeting CBRN and cyber offense misuse domains under Google's Frontier Safety program.
⚒ TOOL SNAPSHOTS
Futuristic tools within AI, no-code, and productivity
🗒️ Meridian - Keeps a running record of your work so updates are easier to write.
🧩 Vendo - Lets customers shape your software around how they actually work.
🧬 Vidaya - Helps turn all those health numbers into clearer next steps.
📚 Skriptr - Keeps AI involved in studying without doing the thinking for you.
🖼 AI ART
Examples of great and trending AI art
Images by xRougeKittenx
ℹ️ ABOUT US
The Intelligent Worker helps you to be more productive at work with AI, automation, no-code, and other technologies.
We like real, practical, and tangible use-cases and hate hand-wavy, theoretical, and abstract concepts that don’t drive real-world outcomes.
Our mission is to empower individuals, boost their productivity, and future-proof their careers.
We read all your comments - please provide your feedback!
Did you like today's email?
What more do you want to see in this newsletter?
