14 LLMs in GAIA, 2024 predictions, and what value is tech without content? All in this week's Content Aware.
Probably the most LLMs in the world...
With a hat-tip to Carlsberg, we're happy to announce the introduction of six new Large Language Models and one new image generation model into Glide's AI Assistant GAIA. They include new models from Cohere, Amazon Titan, Anthropic, and Llama, and raise the number available within the platform to 14 including two image generation models. We think it makes it the widest range offered by any CMS in the world - all safely integrated so you don't have to wrestle with the "who's hoarding my data?" question.
The year ahead
Nieman Lab's annual predictions for the year ahead - by way of analysing the year gone - takes a long look at Google and Facebook traffic scandals, SEO, AI, audience satisfaction, and a host of other factors critical to how our future might look. Excerpt: "With a seemingly limitless number of platforms on which to meet and engage audiences — but still a finite number of hours in the day — teams will need to develop frameworks to understand where to start, stop, and pivot.”
Corbidge comments on... the value of tech without content
What would happen to the mighty social and sharing platforms if there was no content worth looking at or sharing? Probably not much good for the platforms concerned, says our caring sharing Rob.
Good luck tonight
Are you at the British Journalism Awards tonight? More than 800 entries were made to the awards panel, who celebrate the best and most effective journalism over the last year across all channels. Best of luck to all who entered, and come and badger us for a post-event natter in the celebration bar.
State of Ai carries out regular reviews of where the AI industry is at, and has just updated us with its comprehensive 2023 report into the sector and its growing uses, problems, and opportunities. To get up to speed, or delve deeper into industry predictions and investment potential, it's a free download to keep you busy at 163 pages.
Eye spy with my little AI
Not for the first time, news of a site that uses AI to apparently lift content from the articles of others in the same field - except this time it is in the financial news sector and backed by a capital fund. Is it your content?
How to write reviews (according to Google)
While Google has not (yet) started writing its own reviews - although we all remember when it claimed it did - it has produced new advice on how to write them. While it strikes us a bit like the passenger telling the pilot how to fly the plane, what it really shows is Google looks for in reviews to feature in search results.
Riding the SEO rollercoaster
The last ebb of the Google November update is now finished, concluding a nearly month-long roll-out and adding to the turbulence of a stormy SEO autumn overall. If you can stand the pain, RustyBrick has a summary of what's happened since it started. TL;DR: Ground, sky, ground, sky, ground, sky, ground...
So how does the algorithm work?
Meanwhile, here is some once-secret insight into the murky and mysterious calculations which make a Google search result happen, surmised from testimony to the recent Google anti-trust trial.
If you can't beat them
Axel Springer has signed an accord with ChatGPT-maker OpenAI to prioritise and credit its content in search results, as well as to allow use of the publisher's archive to help train past and future models of the AI platform. Other large publishers are likely looking to the same options, as well as to sue for lifting content to train LLMs.
Missed previous editions? Catch-up with them here.