New AI image generator solves text issue

Welcome back to The Daily Update — We’re kicking off the new week with big news from the Air Force and a new AI image generator that solves a problem facing leading models. Here’s the full menu:

🗞️ The Latest in AI: Air Force shows off experimental aircraft run by AI

🤖 AI Tool of the Day: Ideogram — A new AI image generator that can create reliable text within images

📸 Business Spotlight: Alibaba launches two advanced AI models

🚨 AI Roundup: Three quick hits

Read time: 2 minutes

🗞️ The Latest in AI: Air Force shows off experimental aircraft run by AI

The New York Times just reported on an Air Force program that shows how the military is starting to embrace rapidly evolving AI technology. The pilotless XQ-58A Valkyrie experimental aircraft is at the forefront of these efforts.

Key points:

• The XQ-58A Valkyrie is piloted by artificial intelligence agents.

• The aircraft is considered a next-generation drone. The Valkyrie is a prototype for what the Air Force hopes can become a powerful supplement to traditional fighter jets.

• Its mission is to identify and evaluate enemy threats, and then move in for the kill after getting human sign-off.

• The Air Force is planning to build 1,000 to 2,000 of these AI drones for as little as $3 million each (F-35 fighter jets cost $80 million per unit).

Why it matters:

• Gaining and maintaining an edge in AI is a crucial element of an open race with China for technological superiority in national security.

• There is a wide range of specialized roles that these AI aircraft can fill. A significant benefit is that the drones could conduct high-risk missions that are considered too dangerous for a human-piloted plane.

• The U.S. military is confronting concerns about how much autonomy to grant a lethal weapon.

Click here to read the full NYT report (highly recommend checking this out).

🤖 AI Tool of the Day: Ideogram — A new AI image generator that can create reliable text within images

The inability to output verbatim text (or any legible text at all) is a huge weakness in leading AI image generators. A new generative AI startup called Ideogram just launched its text-to-image model that seems to solve this problem.

This new model generates pretty impressive images and is actually capable of generating legible text within your AI images. Ideogram is web-based and works best in Chrome.

Click here to try Ideogram. It’s been automatically opened to the first 1,000 that join and will then go to a waitlist.

📸 Business Spotlight: Alibaba launches two advanced AI models

Alibaba launched Qwen-VL and Qwen-VL-Chat on Friday. The Chinese tech giant says the models can understand images and carry out more complex conversations.

Key points:

• The two new models will be open-source, meaning they are free for research and commercial use.

• Qwen-VL can respond to open-ended queries related to different images and generate picture captions.

• Qwen-VL-Chat caters to more complex interactions, such as answering questions about the locations of certain hospital departments by interpreting an image of a hospital sign.

• The models are built on Alibaba’s large language model called Tongyi Qianwen..

The bottom line: The global race for leadership in AI is heating up. Alibaba’s cloud division is aiming to spark growth as it prepares to go public.

🚨 AI Roundup: Three quick hits

• ChatGPT Update: OpenAI names data-labeling startup Scale AI as its preferred partner for fine-tuning GPT-3.5.

• Tight Rope: Despite cheating fears, schools are repealing ChatGPT bans.

• Big Winner: Nvidia stock continues to impress Wall Street with record-setting surge.

Have questions or thoughts on today’s newsletter? Reply to this email and I’ll get back to you as soon as I can. Thanks for reading The Daily Update!

Jack

(P.S. If you know someone that may find today’s newsletter useful, send it their way!)