- Unfolding AI
- Posts
- It's not out yet | Issue 29
It's not out yet | Issue 29
The Unfolding:ai weekly newsletter about AI for Business Professionals
Paul Bratcher
22 Feb
Hi Everyone,
February is ending with a bit of an announcements and updates splash. This month we have had 2 new approaches to video, 2 new google gemini models (one live, one announced), new voice solutions…. For a moment I thought we were going to have a quieter Q1 compared to last year.
We always like you to get the most value out of the newsletter, in addition to the email version of the newsletter there is a supporting website where you can log on and get to all the previous versions. It’s a trove of resources, especially if you are just subscribed, 28 issues deep,.
I wonder have you considered upgrading to our premium tier, it costs the same as a starbucks a month, and you get over twice the amount of content, including more depth on key topics, key products and how to generate value with AI in your organisation and personal productivity?
We are a growing newsletter, we appreciate every share and recommendation.
Table of Contents
Announcements, but not live
We will look back, and remember the days of 2023, the days before marketing had really got involved in AI, when product announcements were one of two things. Either a research paper, or a product you could use.
Not anymore, it’s easy to get lost in the news of announcements now and then wonder why you can’t actually use them.. yet. So here is a round up of things that will be here in the future, or are in roll out to different geographies/users so you might not have them.
Google Gemini Pro (full features)
The latest version of gemini pro (or is it google one app) is rolling out. You can buy the subscription, and you can use the website version. However, no way to upload files, but you can an image. Which you can ask questions about, or ask it to describe the contents.
Gemini is also really weird still on results, it described one of my images as неизвестная. I have never conversed in russian.. It means ‘artist unknown’, I asked why it suddenly used russian.
You're absolutely right! I definitely could have simply said "artist unknown". Here's why I went down a slightly different path:
Assuming a translation need: I initially assumed you might not be familiar with the word "неизвестен" and wanted a full translation and explanation. It's common for people to send images of objects or text in foreign languages needing clarification.
Providing more context: I wanted to give you a little extra background on the word itself, how it's used, and that it's common to see it in art attribution.
In retrospect, a straightforward "artist unknown" would have been the most efficient and accurate answer. Sometimes I get a little overzealous!
The google app (ios) is supposed to have a tab to allow you to use google advanced, not yet.
Google Gemini 1.5 Advanced
Due for release in ‘February’, an updated version of gemini with a 1.5 Million token context window. This will be a major step forwards, assuming that the current tendency for spurious responses is also trained out of the model. The large context window in effect allows for huge amounts of text (and video apparently) to be processed in prompt. I have seen posts from credible AI specialists, this will probably put pressure on openAI to release something.
Video, Sora
Overshadowing the actual technology releases from stable diffusion openAI SORA is a brand new, not released to everyone, text to video product. It has a new underlying architecture compared to previous solutions. The showreel outputs are remarkable, and also long. The current state of the art is about 4 to 12 seconds, SORA is able to generate 1 minute.
Why is 1 minute important? Its the monetisation duration needed for TikTok.
Microsoft also announcing
To be fair, this is an open source research project, UFO.
“UFO is a UI-Focused dual-agent framework to fulfill user requests on Windows OS by seamlessly navigating and operating within individual or spanning multiple applications.”
It’s a browser extension, which can take control and learn / perform tasks as your own ‘robo ai typing agent’, basically using the web for you.
This is a startup that is (maybe was) doing a similar thing.
Reply