Sometimes the best features are the ones nobody tells you about.
While using the ChatGPT app on my iPhone, I accidentally held my finger on the blue send button a little too long.
Instead of sending my message, a hidden menu appeared.
It lets you choose how much ChatGPT should “think” before answering.
If you’ve never seen it before, you’re probably not alone.
Long-pressing the send button opens three intelligence options:
• Instant for the fastest response
• Medium for a balance of speed and depth
• High for more detailed reasoning before ChatGPT replies
These options let you choose how much time the AI spends processing your request.
For quick questions, Instant is usually enough.
For coding, writing, research, or complex problems, High can produce much more thorough answers, although you’ll wait a little longer.
Using the feature is simple:
Type your prompt.
Long-press the blue send button.
Choose Instant, Medium, or High.
Send your message.
Unlike the web version, where these options are easier to find, the iPhone app doesn’t make their existence obvious.
It’s one of those hidden shortcuts that many users may never discover.
There’s one limitation.
The feature is currently available only for paid ChatGPT subscribers.
If you’re using the free version of ChatGPT, the hidden menu won’t appear.
Android doesn’t offer the exact same shortcut.
Instead, users can access similar options by changing the model they’re using.
Depending on the version of the app, you can switch between faster and more reasoning-focused models through the model selection menu.
The experience isn’t identical to iPhone, but it offers a similar level of control.
One of the strengths of modern AI is that not every question needs the same amount of processing.
Sometimes you just want a quick answer.
Other times you want the model to spend more time reasoning through a complex problem.
This hidden menu makes it easy to choose the right balance for each conversation.
It’s a small feature, but once you know it’s there, you’ll probably find yourself using it far more often than you expected.
A journalist shared how an AI coding agent accidentally generated nearly $500 in OpenAI token charges after getting stuck in a loop while working on a project. The issue went unnoticed for days because the agent was running remotely, repeatedly consuming tokens as OpenAI automatically added $5 credits to the account. The experience highlights how AI agents can quickly rack up unexpected costs when given too much autonomy.
The incident underscores a growing challenge for both individuals and businesses using AI tools: managing token spending. While AI coding assistants can significantly boost productivity, users are encouraged to set spending limits, monitor long-running tasks, and avoid giving agents unrestricted access to payment methods until the technology becomes more reliable.
Elon Musk claimed Grok 4.5 could match Claude Opus-level performance while running faster and at a lower cost. Now, an independent benchmark from Artificial Analysis appears to support that claim. On the AutomationBench-AA benchmark, which measures how well AI agents complete real-world tasks across apps like Gmail, Slack, Salesforce, and HubSpot, Grok 4.5 ranked first, outperforming both Claude Fable 5 and Claude Opus 4.8.
The biggest advantage was efficiency. Grok 4.5 completed tasks for just $0.34 each, compared with $1.35 for Fable 5 and $1.46 for Opus 4.8, largely because it used far fewer tokens. However, the benchmark also found that Grok violated guardrails more often than its closest competitors, suggesting companies deploying AI agents in sensitive environments may still need to balance speed and cost against reliability and safety.
AccountKiller provides guidance on deleting accounts from different websites and platforms.
It is useful when you want to clean your digital footprint or remove old accounts you no longer use.
That’s it for today.
The AI race is accelerating - new breakthroughs, new tools, and new possibilities are appearing faster than ever.
The biggest risk isn’t AI replacing you. It’s someone using AI better than you.
Until next time: stay curious, stay ahead, and keep exploring the future of intelligence.
No posts
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.