The Anthropic vs OpenAI competition is working in your favor.
These roundups have been more technical lately. That's where the action is. What you shouldn't miss is that even if you don't use this stuff it's still helping you.
Let's breakdown what's going on:Power users have been heavy on Claude for the past year. Mainly because of Claude Code (AI working in your files and spinning up agents for coding and non-coding tasks). Companies noticed and started moving money from ChatGPT to Claude.
OpenAI panicked, shutting down random side projects in March and throwing everything at coding and enterprise.
Last month they shipped GPT-5.5. It's good, so power users flipped back. Meanwhile, everyday users are discovering Claude.
These swings are rough on the labs, but they're great for the rest of us. We get better models, faster releases, higher usage limits, lower prices, and new agent features.
Even if you're stuck on one platform for work, or none of this touches you directly, you don't have to switch tools to benefit.
As long as the labs are scared of each other, you win.
-Riley
|
This week's top updates
|
|
|
|
FEATURE
May 14
Start, review, and steer Codex tasks from iOS or Android while your computer keeps doing the work.
|
-
Best for: Codex users who want to kick off or review tasks away from their laptop.
-
When to use it: Trigger work, approve next steps, or monitor an in-flight task from your phone.
-
Who benefits: Developers running Codex on a laptop, Mac mini, or devbox.
|
|
|
FEATURE
May 11
OpenAI's product for cyber defense teams, combining frontier models, Codex, and security partners.
|
-
Best for: security teams automating threat detection, validation, and response.
-
When to use it: you're a defender trying to keep pace with attackers' speed.
-
Who benefits: SOC analysts and security engineers who currently triage alerts manually.
|
|
|
|
|
|
MODEL
May 12
Google's low-cost, low-latency Gemini model is now generally available, with response times under two seconds for full replies.
|
-
Best for: high-volume tasks where response speed and per-call cost dominate model choice.
-
When to use it: you're building agents, classifiers, or tool-callers that run at scale.
-
Who benefits: developers running real-time agents, customer support automations, or IDE assistants.
|
|
|
FEATURE
May 12
New Gemini Intelligence features rolling to Samsung Galaxy S26 and Pixel 10 first: Rambler, Create My Widget, multi-step app automation, intelligent autofill, and Chrome assistance.
|
-
Best for: Hands-off mobile workflows that span multiple apps like booking, shopping, or comparison research.
-
When to use it: Offload multi-step phone tasks like comparing prices across tabs, drafting messages by voice, or filling long forms.
-
Who benefits: Android users on Samsung Galaxy S26 and Pixel 10 first, broader rollout later in 2026.
|
|
|
|
|
|
FEATURE
May 14
Richer Similarweb web intelligence inside Manus for growth, traffic-source, and keyword analysis.
|
-
Best for: Growth, SEO, and competitive-research workflows in Manus.
-
When to use it: Need to understand what drives traffic on a site, not just total volume.
-
Who benefits: Manus Pro users who run market and competitor research inside the agent.
|
|
|
|
See all updates
|
Learn AI by Doing
|
 |
|
People getting the most from AI are using this system, where AI finds the info it needs and gets better the more they use it. Build yours now, no experience needed.
|
|
Start this module
|
|
 |
|
Turn a task you do in your browser regularly into a one-click Gemini-managed workflow.
|
|
Start this module
|
|
|
| See more modules → |
|
