Product Hunt Weekly Digest
May 6th, 2018

Atari’s new console revealed 🕹️
Atari Box 🎮
Last year, Atari released a teaser of their newest, secret console. The Product Hunt community was quick to speculate:

Atari Europe and I literally share the same floor in the same building! They're so secretive. My guess is retro console."Pierre-Marie Galite

Pierre-Marie was right. The AtariVCS was announced today, with pre-orders starting at the end of the month:

🕹️ 100+ classic games including Asteroids, Centipede, and Breakout
🖥️ Supports 4K monitors at 60 FPS, comparable to an Xbox One
💰 $200 for the basic model, with a Collector's Edition coming soon

Atari has seen better times: they've been acquired a number of times and only recently emerged from bankruptcy. It’s been over 25 years since Atari launched a new console.

Meanwhile, Nintendo has succeeded by leveraging nostalgia. Pokémon Go exploded two summers ago as fans swarmed parks worldwide to catch 'em all. They also released a mini replica of the beloved NES. Their stock price has tripled in less than 2 years.

We can only hope that Atari will follow in Nintendo's footsteps. Unlike like the Nintendo Switch or Xbox One, the AtariVCS will be crowdfunded on Indiegogo. Anyone can sign up to buy one here.

If you can’t wait for your AtariVCS, you can play Defender, Joust, Paperboy, and 9,000 other retro games in your browser for free, legally.
Best landing page. Ever.

Building machine learning models is difficult. You'll have to learn Python or R, then understand the various open-source platforms like Google's Tensorflow, format your data correctly, and hope your model works.

"A holy shit moment in tech from our company Lobe. Full AI capabilities that you can put in your app but without having to write any of your own AI code! Watch the demo. It's bananas." – Chris Sacca

Lobe just launched to make building machine learning models easy:
1. Label your training set with your desired results.
2. Watch as the algorithm builds itself.
3. Ship it. Export your model into Google's TensorFlow, Apple's CoreML, or directly into your app on iOS or Android.

Teams are already building incredible projects:
😳 A tool that distinguishes between skin cancer and moles
🌱 An app that identifies California plants by photo
🏘️ A model that identifies architecture types from 3D models

Lobe’s landing page might be the best product landing page we’ve seen since Muzzle. 👌