Hey {{first_name|Conscious Church Fam}}
We’ve got models solving Olympiad-level maths, decoding ancient Roman inscriptions, and designing cancer treatments in a matter of weeks. OpenAI’s even teasing GPT-5 for August.
In today's recap:
🥇 Google Gemini officially wins gold at the Maths Olympiad
🤔 OpenAI says “same here” but didn’t go through IMO
🏛 AI helps restore Roman Empire inscriptions
🧬 AI designs cancer-targeting proteins in weeks
📅 GPT-5 is launching in August
Let’s dive in 👇
💭 Josh’s Musings
I’m sharing at a youth conference this coming weekend about Creativity in a World of AI.
And... I’ve been pondering that thought for a while maybe even a few years.
We’re surrounded by smart tools. Stuff that can generate images, write essays, edit videos—all in seconds. It’s wild.
But I think the danger is we (me anyway) start thinking creativity is just output. Just volume. Just doing more, faster.
It’s not.
Creativity is how YOU see.
It’s how you respond to life on your journey.
AI can make cool stuff, but it can’t carry conviction. It can’t sense what a room needs. It can’t bring breakthrough.
That’s the edge we’ve got.
Not better tech. A better Source.
I keep coming back to this:
If the Spirit of God lives in us, then the way we make, lead, speak, design, it should look different.
You don’t need to copy what’s trending. You don’t need to force it.
You just need to stay close to Him. And start creating from there.
Because yeah, the tech’s getting wild, and to some extent, everyone is now their own creative director
But people full of the Spirit, showing up with what God’s put in them, and with a good eye and taste…that will always be more powerful.
🙌 Stay Curious, Stay Conscious, Stay Wild
Josh
LATEST NEWS
Image Source: Google DeepMind
Recap:
Google DeepMind's latest version of Gemini has scored 35/42 on the International Mathematical Olympiad—enough to earn a gold medal. This time, they went through the official IMO process to verify it.
The Details:
Tackled six complex maths problems in a timed 4.5-hour test
Solved five, placing it at gold standard level
Fully natural language model, no special hacks or translations
Officially graded by IMO coordinators using standard criteria
Conscious Take:
This milestone shows how fast AI is moving toward advanced reasoning. It’s not just solving basic prompts anymore, it’s handling elite-level problem-solving in real time ‘under pressure’.
Image Source: The Conscious Church / ChatGPT
Recap:
OpenAI said their new model also hit gold-level performance (35/42), but their version wasn’t submitted for official grading. Instead, former IMO medalists reviewed the results.
The Details:
Same problems, same time constraints
Graded by three past Olympiad winners, not the official IMO team
Model name not disclosed
DeepMind publicly pushed back, saying official verification matters
Conscious Take:
These kinds of benchmarks mean a lot to the AI world. But as models compete to outdo one another, the methods behind the claims matter as much as the results. Trust and transparency will only get more important.
Image Source: Google DeepMind
Recap:
Google DeepMind launched a tool called Aeneas, designed to help historians restore damaged Latin inscriptions from the Roman Empire. It analyses fragments and suggests restorations, time periods, and locations.
The Details:
Trained on 176,000 Latin inscriptions
Attributes text to specific Roman provinces with 72% accuracy
Can date inscriptions within a 13-year range
Historians said it improved their confidence by 44%
Open-source and usable for other ancient languages
Conscious Take:
It’s a powerful example of how AI can serve restoration, not just invention. Tools like this are giving historians new ways to piece together stories that were nearly lost.
Image Source: Science 2025
Recap:
Scientists in Denmark have built an AI platform that can rapidly design “minibinder” proteins to help immune cells hunt and destroy cancer cells—especially personalised to each patient.
The Details:
Targets cancers like melanoma with tailored protein designs
Uses AlphaFold2 for structure prediction
Includes built-in safety screening before lab testing
Designs go from idea to lab-ready in weeks
Conscious Take:
We’re seeing AI compress time in extraordinary ways. What used to take years in medical R&D is now happening in weeks. It’s a glimpse of how AI might transform treatment timelines, and save lives in the process. That’s some good news!
Image Source: Theo Von YouTube
Recap:
OpenAI says GPT-5 will launch in August. It combines high-level reasoning and language into one model, removing the need to switch between tools depending on the task.
The Details:
Altman said testing it made him feel “useless”
Not the same model used for the IMO test
Public release is coming soon, with an open-weight model also on the way
Goal is to simplify workflows and improve reasoning at scale
Conscious Take:
GPT-5 is being positioned as a major leap, faster, smarter, and more integrated. The details are still under wraps, but expect it to raise the bar for what these tools can do across every area of life and work.
“For the Lord gives wisdom; from His mouth come knowledge and understanding.” – Proverbs 2:6
That's all for now
Stay conscious,
Josh
P.S. If you liked this then please forward it on to someone you think would enjoy it. And if someone forwarded you this and you liked it, you can sign up here.
