GM! Your AI-obsessed penguin here, waddling through the news so you don't have to.

Here’s what I have for today:

1. AI Models Are Getting Expensive AF (But There's a Plot Twist)

2. OpenAI's About to Get a $40B Cash Injection

3. Meet o3-mini: The New Model That Won't Break Your Bank

(Estimated reading time: 3 minutes 45 seconds)

Not subscribed yet? Stay in the loop with AI daily by reading Get Into AI for five minutes.

Three main stories for the day.

1/ TRAINING COSTS ARE WILD, Y'ALL

A fresh cost comparison of AI model training just dropped, and... whistles. The upcoming 100K H100 Scale Models are estimated to cost around $400M to train. For perspective, that's about 80x more than training DeepSeek V3 ($5M~ which is just the training cost).

This massive cost gap explains why only a handful of companies can play in the big leagues.

The AI arms race is getting spicier (and pricier) than a ghost pepper challenge, but DeepSeek's success suggests there might be more efficient paths to AI excellence.

2/ OPENAI'S MEGA FUNDING ROUND

OpenAI is in talks for a casual $40B funding round, potentially pushing its valuation to $340B. SoftBank's Masayoshi Son is ready to lead with a $15-25B check, potentially overtaking Microsoft as OpenAI's sugar daddy.

This isn't just about the money – it's about building AI infrastructure through Project Stargate (a joint venture with SoftBank and Oracle).

Plus, they're gearing up for intense competition with xAI, Microsoft, Google, and that new kid on the block, DeepSeek.

OpenAI's playing 4D chess here – securing both the bag and the infrastructure while DeepSeek's making waves in their backyard.

Even Sam Altman admitted DeepSeek is "clearly a great model." Game on!

3/ O3-MINI: OPENAI'S NEW POCKET ROCKET

OpenAI just dropped o3-mini, their latest reasoning model, available on both ChatGPT and API. Think of it as the Tesla Model 3 of AI – smaller, faster, but still packing a punch in STEM capabilities.

This is OpenAI's first small reasoning model with developer-friendly features like function calling and structured outputs. Plus, free users can finally play with a reasoning model (this is huge!).

OpenAI's making AI more accessible while maintaining quality. With 39% fewer major errors than its predecessor and triple the rate limits for Plus users (150 messages/day), o3-mini is bringing more bang for less buck.

🐤Tweet of the week

Is this going to be the future of education? Good food for thought!

Catch you tomorrow! ✌️

That’s it for this week, folks! If you want more, be sure to follow our Twitter (@BarunBuilds)

🤝 Share Get Into AI with your friends!