🐺 SuperGrok. Super slow. Super annoying?

 📰 What’s Trending

OpenAI wants to hire you and replace recruiters while it’s at it
OpenAI just launched Hire, a recruiting platform that screens applicants using its AI models. It’s aiming squarely at LinkedIn, with built-in assessments and candidate summaries. This could change how SMBs and startups hire—faster screening, less résumé fluff.
🔗 OpenAI announces AI hiring platform to take on LinkedIn

These Gen Z founders just raised $5M to analyze your online habits
Ages 20 and 22, YC-backed founders of Optic are building “vision AI” to watch how users interact with websites. It's Google Analytics meets screen recording, but with machine learning doing the analysis. Anyone running CRO or growth experiments should keep an eye on this.
🔗 These 20- and 22-year-olds raised $5M to study online behavior

Anthropic pulls in $13B, now worth $183B
Claude’s parent company just raised a monster Series F….$13B from Amazon, Google, and others! That’s more than OpenAI’s entire 2023 revenue. There’s truly no ceiling on foundational model bets (yet), and Claude’s likely getting more business-focused features soon.
🔗 Anthropic raises $13B at $183B valuation

PayPal and Venmo start handing out Comet invites
Comet, the AI-powered (sorry) shopping assistant from ex-Amazon folks, is now integrated into PayPal and Venmo. It scans your spending and suggests products or savings. If you’re in ecommerce or fintech, this could shape how people browse and buy.
🔗 PayPal and Venmo are now giving out Comet invites

Scale AI sues a rival for allegedly poaching clients
Scale is taking Mercor and a former employee to court, accusing them of trying to steal major contracts and confidential data. Yes, it’s messy but also a reminder that the AI gold rush is cutthroat, and customer relationships are the new oil.
🔗 Scale AI sues rival Mercor over client poaching

This AI-generated Will Smith video has the internet confused
A weird, deepfake-style video of Will Smith in a crowd has gone viral and no one can agree if it’s real (it’s not). “AI or Not” is becoming a real genre now. Marketers, be ready: brand trust is going to hinge on proving authenticity.
🔗 Will Smith’s crowd video sparks AI confusion

Lovables CEO shrugs off AI competition
The founder of vibe-coding app Lovables says he’s not sweating OpenAI’s entry into the space. He knows users want emotional nuance and a brand that gets them, not just productivity. It’s a lesson in brand identity: standing out might matter more than scaling fast, especially in the long term.
🔗 Lovables CEO isn’t too worried about AI vibe competition

🧠 To Think or Not to Think

How This All Started

Among all the uproar around ChatGPT 5.0, not only did the model not perform as well as people had hoped, but they sneakily added this "thinking" feature. And it's not just GPT-5 - Grok did the same thing, now calling itself SuperGrok to disguise the fact that its default mode is thinking mode, which takes substantially longer to process and answer questions.

To me, this feels like a way to cut costs and experiment with whether people mind longer wait times in return for their servers getting more breathing room. They basically took the fun out of hitting Deep Research when you actually want a deep dive on something specific. Now it's been replaced with thinking mode, which I guess makes sense because sometimes the models don't take their time.

Here's What I Think About It

Look, I found one main pro among all the cons: when I'm asking a model to do something and it's just not following my directions, instead of having to say "stop, slow down, let's explain the problem and why you're having issues," I can throw it into think mode. Sometimes it works, sometimes I still need to switch LLMs completely.

But I don't love this function. There are three options - Instant, Thinking, or Auto. It's set to auto by default, and when ChatGPT 5 first dropped, auto really felt like it meant defaulting to thinking. I'd ask a simple question and it would take 3 minutes to spit out an answer when really all I needed was speed.

We got so used to instantaneous responses from these models. They could think for maybe 10 seconds and generate 30-page documents. When the thinking process takes longer than the actual output, that's when my jaw was really dropping about AI. Since most things focus on speed and efficiency (literally the value propositions that businesses compete on), when AI slows down it feels backwards.

What I'm Seeing Across Models

The difference between GPT-4 and GPT-5 doesn't seem that noticeable to me, but maybe I haven't used it enough. The only thing I really notice is having to annoyingly switch to Instant mode.

For Grok, which is the least used LLM in my tech stack out of the big four (Gemini, Claude, GPT, Grok), it seems to default to think mode immediately. SuperGrok might handle large research tasks well for academics and major corporations, but for the average user it just made things more inefficient.

I typically only use Grok for social media strategy, trend analysis, current events content, and real-time market insights anyway. If I want deep research, I'll probably just use Manus since I like seeing the AI agent working in real-time and being able to interact with it while it's working. It definitely takes longer, but that's what I'm used to with Manus. When you drastically change the time it takes other models to answer, it makes the user experience worse and the threshold to see which model is actually better much thinner.

The Reality Check

I've seen research showing people seem to like GPT-5 and SuperGrok - they definitely have better grades compared to their previous models. But it's interesting how a large group of people are deterred by the long waiting time since we're so used to instant answers and gratification.

If this was how AI and LLMs had started off, nobody would care or make a fuss. But since it's so much longer than the norm, it feels like we took a step back.

What's Next and Why This Matters

Deep Research function becoming thinking mode? Fine. But let's not auto-default to think mode and make it a core function. The main benefit and value prop of these AI platforms is getting incredible answers and content in minutes.

When speed and efficiency are what businesses compete on across every industry, slowing down AI without clear user benefit feels like missing the point entirely. The most sophisticated model in the world doesn't matter if people stop wanting to use it because the experience feels clunky.

💡Smart Moves

You’re invited to the world's largest email marketing conference.

Become an email marketing guru at the GURU conference. It’s two days full of all things email marketing. Learn more about newsletters. deliverability, design trends, AI, and what NOT to do with email.

What you can expect:

• Keynote Speakers: Nicole Kidman, Amy Porterfield & more!

• The latest digital trends in email marketing & how to increase performance.

• Networking opportunities - each day!

• Dj’s, dance contests (judged by Lance Bass, yes for real), breaking world records & MORE!

Spots are limited. It’s VIRTUAL. It’s FREE. It’s time to become an email marketing GURU. Join 25,000+ marketers on November 6th & 7th. Don’t miss out!

YES! I want to register!

🤣 Byte-Sized Laughs

💎 Data Gem

72% of marketers say AI has already changed how they create content.

According to a 2025 HubSpot survey, nearly three-quarters of marketers report shifting their workflows around AI tools. That means faster draft cycles, automated research, and more testing.