In partnership with
Tech moves fast, but you're still playing catch-up?
That's exactly why 200K+ engineers working at Google, Meta, and Apple read The Code twice a week.
Here's what you get:
Curated tech news that shapes your career - Filtered from thousands of sources so you know what's coming 6 months early.
Practical resources you can use immediately - Real tutorials and tools that solve actual engineering problems.
Research papers and insights decoded - We break down complex tech so you understand what matters.
All delivered twice a week in just 2 short emails.
Good morning!
Anthropic reveals a secret AI that finds major security flaws but wonβt be released. Plus, a new model builds full apps on its own and beats top rivals.
In todays email:
Daily Update
Todayβs Highlights
YouTube
Today Trend
Social Media
Read time: 6 min
DAILY UPDATE
Anthropic revealed a secret, powerful AI called Mythos that found major security flaws and will not be released publicly.
Anthropic has launched Project Glasswing, a cybersecurity group with big tech partners like AWS, Apple, Google, Microsoft, and Nvidia. At the center is Claude Mythos Preview, a highly advanced AI model with very strong abilities.
Mythos discovered thousands of security issues, including bugs hidden for over 27 years across major systems and browsers.
It shows huge performance gains in coding, reasoning, and other key areas compared to top AI models.
The model will not be public, and access is limited to a small group with $100M in support for security work.
This shows how powerful hidden AI models have become. Anthropic is being careful, choosing to focus on safety and cybersecurity first before releasing anything at this level to the public.
Continue Readingβ¦
TODAYβS HIGHLIGHT
Chinese AI model GLM-5.1 beats top rivals in coding and runs tasks for hours
Chinese AI lab Z AI has released GLM-5.1, a new open-source coding model that matches and even beats top competitors. It is built to handle long, complex tasks, working continuously for up to 8 hours without help.
GLM-5.1 scored 58.4 on SWE-Bench Pro, beating GPT-5.4 and Opus 4.6, taking the top spot
The model can stay effective on long agent tasks, solving complex problems over extended time
In testing, it built a full Linux desktop web app on its own, including tools like a file browser and terminal
This shows that open-source AI is catching up fast with top models. It also highlights a shift toward AI that can handle long, real-world tasks, which may become more important than just scaling size in the future.
Continue Readingβ¦
YOUTUBE
Table Tennis AI vs. strangers
TODAY TREND
Velo
Share anything as video messages Promoted
Flint
Launch on-brand pages for every campaign, ad, and prospect. Promoted
Google Chrome Vertical Tabs
Chrome now supports vertical tabs and immersive reading mode Promoted
LookAway 2
The Mac app your body thanks you for
MindsDB Anton
Business intelligence that doesn't just answer β it acts
SOCIAL MEDIA
Thatβs it for today!
Before you go weβd love to know what you thought of today's newsletter to help us improve the experience for you.
