Yes Im starting to share your sentiment on AiM @DrPete. I think it needs to evolve further. I have been jumping in to quickly on some of these fledgling companies and having to do the long hold which isnt my style unless the compoany is well established. Ill keep an eye on it from the charting prospective however I will probably back off and increase the thresholds / price targets I use for my alerts. Then I will revisit the charts and commentry on Strawman then. Thanks also to @Arizona & @Wini for the content that helped me make this decision

@Saiton I really want to see AIM do well. I would really like to be a shareholder. I hold out hope.

Tony buying in more and his ownership is great to see. I do however, remain cautious.

I am intrigued that there don't appear to be any great examples of the companies products (Lexi etc) in action. The company website has flash photos, diagrams and explanations, but it is not easy to find good examples of video showing how this all works. I would have thought such examples would be a no brainer. I'm a visual kinda fella.

I may have missed something. Not looked hard enough. If someone has a/some good examples, please let me know. You might convince me to get on board.

I must reiterate that the discussion here on SM has been incredibly valuable. Thanks all

Held on SM

@Arizona You can tune into Channel 7 or 9, turn on the captions and you'll able to see LEXI 3.0 in action. The best use case is live TV - such as the Today show - where all the content is live and one can observe the live latency.

I don't watch any free-to-air TV, but I did watch the last Olympics on Channel 9, and couldn't fault the live captioning. I also tuned into the Tyson vs Paul fight on Netflix just to catch the captioning - also couldn't fault it :D.

@mushroompanda This is great. I should have picked this up!! I have seen that graphic a number of times now.

I am not a TV watcher.

That has to change, I have get out there and watch some TV. Its research.

Thanks @mushroompanda much appreciated.

Yes I also must agree that Strawman is becoming helpful to me also. I see both sides evryone has put forward. I will keep an eye for now and wait to see more of definite turn around on sentiment by the market before I get involved. I have been jumping on board to early on some stocks and need to pull myself back in check. After all the market drives the $. Thankyou everyone for the info/views brought to my attention to help me make better decisions on my part.

@mushroompanda, do we know whether these captions are AI-only generated or still involve human input? My concern is their AI-only solution. If their AI-only solution does not exceed human captioning it will mean the transition to higher margin technology solutions is slowed. Which will create significant headwind for revenue and margin growth and achieving Tony's aspirational targets.

A human captioner is hundreds of dollars per hour. If they were using humans it would be showing up in ballooning employee costs given their stated use of Lexi minutes.

The transition to AI is part accuracy and part costing 80% less than using a human. Broadcasters will not quibble over >95% accuracy when the difference in cost is orders of magnitude.

@DrPete I don't know for sure, but I expect Channel 7 and 9 to have been migrated completely to AI-only by now. The company has flagged it'll be 100% complete by Dec 2025.

I know the Olympics and the Tyson fight were done by LEXI because that's what the company had flagged publicly.

For something we know is certainly done by LEXI 3.0 is the AWS events. You can see the live captioning at certain camera angles: https://www.youtube.com/watch?v=LY7m5LQliAo

AWS has wrote about it in detail here: https://www.linkedin.com/pulse/training-ai-make-live-experiences-more-accessible-taylor-wilkins-gep2c/?trackingId=IuMtFII9SZapBXKGvi2oQA%3D%3D

@mushroompanda many thanks for the link to AWS article. This was really useful as it provides another external, non-AIM, technical view of how LEXI Voice is wired up and baked into the overall AWS live feeds.

Here is how I see the pieces described in the article against the AIM ecosystem slide - colour coded the relevant text/arrow to the component on the AIM slide, so this picture is best read alongside the article text:

• Live Stream - this is straightforward - that's the raw media input, LIVE
• Encoders - the photo is a AV610 encoder, which from a Google, is an EEG physical encoder (as compared to more recent Falcon and Alta, software-based encoders) - good to see a picture of the encoders, with the AIIM logo on it!
• Customer Automation Systems - the most interesting bit for me was the bit about the Topic Models and Contextual Training - that would fall under the “Customer Automation Systems” - the unique-to-each-customer contextual information that is fed into Lexi Cloud, and then used by the entire LEXI ecosystem to make sense/insert/replace AWS-centric jargon into something LEXI Voice would not ordinarily recognise from pronunciation alone, but would now be correctly captioned

When broken down, it clearly shows how the LEXI ecosystem “orchestrates”, technically brings together/integrate the different pieces REAL TIME during a LIVE feed, to produce the translations and captioning.

Note also the AI/ML models are occurring inside the AWS systems, used to make sense/tech the Topic Models mentioned - AIM is thus not competing with any AI/LLM’s, but both AIM and the Customer are USERS of LLM’s

The more I read articles like these, the clearer the interplay between the Encoders/LEXI Cloud and Customer Systems becomes for me!

Last point, this was written in July 2024, 1 full year ago - things would have improved significantly since then, given how fast anything technical and AI moves these days.

Discl: Held IRL and in SM