Google's Nano Banana is incredible.

How to turn any photo into a 3D AI figure:

1. Open Gemini and select 2.5 Flash
2. Upload your image
3. Click “Create images”

Use this exact prompt (in comments): Full prompt:

“Create a 1/7 scale commercialized figure of the character in the photo. The style should be realistic, with clearly defined features, and placed in a real-world environment. The figure should be positioned on a computer desk, standing on a round transparent acrylic base with no text. On the computer screen, display the Adobe Illustrator modelling process of this figure. Next to the screen, place a BANDAI-style toy packaging box printed with the original photo.”

OpenAI just published their official prompting guide for GPT-5.

Master these 6 critical prompting techniques: 1. Be precise and avoid conflicting information
↳ GPT-5 models are significantly better at instruction following
↳ But they can struggle with vague or conflicting instructions
↳ Clear, specific prompts = better results

I just fell for one of the best scams I've seen on X.

Here's what you need to know (and how to avoid it): 1. The intro

On Thursday last week, I received a message from "Marthe" at TechCrunch offering to do a podcast conversation about AI.

I do a lot of media collaborations, so I thought this could be a nice opportunity.

Her profile looked normal with TechCrunch article reposts.

Google just announced Med-Gemini for medical tasks.

It beats GPT-4 on all benchmarks: Med-Gemini is a family of Gemini models fine-tuned for medical tasks.

The MedQA-USMLE benchmark is a dataset of USMLE-style multiple-choice questions used to evaluate the medical knowledge and reasoning capabilities of AI systems.

I just launched my course on Understanding Prompt Engineering.

My 10 favourite strategies to get the best results when prompting ChatGPT: 1. Clarity

• Include the relevant context to reduce ambiguity.
• Example: “I'm an 8th-grade math teacher preparing to teach trigonometry.”

The most popular motorsport in the world.

Formula 1.

In 2022, it had a cumulative TV audience of 1.54 billion.

Here's how AI and data are disrupting the sport forever: At the 2022 Canadian Grand Prix, Fernando Alonso was starting P2 on the grid.

This was his best qualifying result in a decade.

Before the race began, he was asked:

"I know you’ve joked about passing Max into turn 1. Is that a realistic possibility?"

In his essay, "Why AI Will Save the World," Marc Andreessen presents the argument that AI could do more than make our lives better—it could quite conceivably save the world.

Here's why AI has the potential to revolutionise life on Earth: Marc Andreessen is the Co-Founder and general partner of the VC firm a16z.

He refutes fears surrounding AI and presents an action plan for the West to achieve AI dominance.

Andreessen believes AI is just a tool—like fire and rocks—that can be used for good or bad.

"There will be no programmers in 5 years."

In March, Emad Mostaque, the Founder & CEO of Stability AI sat down with Peter Diamandis to discuss how AI and humanity will intertwine.

My 10 favourite takeaways: 1/ We have figured out how to make humans scale

You now don’t need as many humans.

"Using GPT-4 is like using a really talented analyst with a bad memory."

When we fix the memory component, this could mean huge disruption or huge abundance.

OpenAI just released its Code interpreter to all Plus users.

You now have access to a junior programmer at the speed of your fingertips.

Here's everything you need to know: Code interpreter is a new experimental feature. It can:

• Use Python
• Access files you've uploaded
• Download the results of your work

Let's explore some use cases.

9 years ago, Peter Thiel hosted an AMA on Reddit.

It's full of gems on AI, technology and startups.

Here are my favourites: 1/ Advice for entrepreneurs today

Harvard’s new computer science teacher is a chatbot.

Students enrolled in the university’s flagship CS50 course will have access to the AI teacher from September.

Here's the breakdown: Harvard is adopting AI as an official learning tool for its introductory course on computer science.

CS50 professor David Malan wrote in an email statement students can use the AI to:

• Answer individual questions
• Help find bugs in their code
• Explain error messages

Last week McKinsey released their generative AI report.

Here are the 10 key takeaways everyone must know: 1/ Value potential by industry

Generative AI has the potential to generate $4.4 trillion in value across industries.

• $460 billion in high tech. Predominant factor: software engineering
• $390 billion in retail. Predominant factor: marketing and sales

There are 1000+ new AI products released every week.

Get ahead of 95% of people with these 6 proven strategies: 1. Focus on solving a problem for a specific niche

Find a burning problem.

It's got to be red hot.

I see a massive opportunity in domain-specific niches vs general-purpose tools.

• Calendar assistants
• Data analysts
• Travel agents

Excel is finished.

Imagine having a personal AI analyst for your data.

This tool changes everything: Today Rows launched its AI analyst.

It's able to:

• Answers questions about your data
• Summarise any dataset
• Run deep dive analysis

As a finance major, this is the biggest AI update to spreadsheets I've seen.

In the next 10 years, humanity will experience as much progress as we have had in the past 100 years.

What will be the generation-defining technology breakthroughs?

I'll start things off: 1/ Generative AI

Bloomberg Intelligence expects the market for Generative AI to reach $1.3 trillion in 2032.

2022's market: only $40 billion.

It will disrupt everything from customer service to banking.

Apple just unveiled a revolutionary product—Apple Vision Pro.

What does this mean for the future of AI?

Let’s explore the possibilities: Apple Vision Pro is a spatial computing device with visionOS.

Apple's first-ever spatial operating system.

You can:

• Navigate with your eyes
• Tap to select
• Flick to scroll
• Use your voice to dictate

Let's explore how AI can enhance this experience.

Note-taking sucks.

Here's how AI will transform your notes into a second brain: The way we take notes is fundamentally broken.

You learn something → write it down → forget about it.

Luckily GPT-4 has the potential to turn our notes into an AI-powered second brain.

Here are 3 ways this could look.

OpenAI just released their article on the 'Governance of superintelligence'.

"Within the next ten years, AI systems will exceed expert skill level in most domains."

Here's everything you need to know: 1/ The timeline is near

AI is evolving at an exponential pace.

In the span of the next decade, AI systems will surpass human expertise in most fields.

We're not talking about science fiction, but a tangible future that's almost within our grasp.

You don't need "10 mindblowing prompts for ChatGPT".

The best results come from being specific.

Use the persona-problem-solution framework instead: 1/ Persona

Here you want to define the persona of the language model.

• Advisor to Ray Dalio
• Expert problem solver for leadership tasks
• Detailed prompt writer for large language models

Now ChatGPT has context for the nature of the problem it will solve.

Photoshop is finished.

Imagine being able to perfect any picture exactly how you want it.

Now it's as simple as drag and drop with this AI: Researchers from Max Planck Institute for Informatics, MIT and Google have developed DragGAN.

It lets you 'drag' any part of a picture to exactly where you want it.

Geoffrey Hinton is the godfather of AI.

He said that AI is a more urgent risk to humanity than climate change.

Here's what I found: Geoffrey Hinton joined Google in 2013 after selling his AI company.

In 2018, he received the Turing Award—the computing world’s equivalent of the Nobel Prize.

Only recently he left Google after more than a decade.

His reason?

Fears about the rapid development of AI.