The Presentation24 Aug 2025 22:07
OK. Other than the claiming to have commercialized something[s] that has not been commercialized I kind of get Mr Nag's reason for being and things do get better... but.
As suggested fellows get exposure to AI by working for 'free'. You do actually do final, and other, year projects for 'free' whilst still paying for your tuition costs. MSc's and, in particular, PHd's need external funding. I assume Mr Nag is dialing in the 'free' stuff whilst acting, along with others, as Mentor and possibly, probably, paying for the LLM tokens.
Great. It's experience. I also got the impression, in the later examples, that was in part a focus on the commercialization of product. Use cases, Costs, UPS's. As such it is rounded.
However, the but. Across all of the projects it seemed to me that fundamentally third party tools, tokens paid for by Mr Nag and R42, were being glued together with interfaces to get the job done. Yes that works to make use more transparent to the end user but eventually that work will be snaffled or reproduced by others like what happened with Linux.
Ultimately non of what I saw was a Market Killer or if you prefer Unicorn. Again it is all use of third party tools with nothing unique or representative of IP for R42 or their fellows.
For me Author42 froze towards the end. It and a couple of the others had what might be thought of as a USP in terms of work flow and in this case integration with a wider market or rather the mechanisms of that market. Again nothing that might not be copied. Think Dragons Den. However 'first to market' can apply.
Unfortunately this one gets an extra but. I've poked about with ChatGPT 4, one of a number of models mentioned in the presentation, to see what it might be like to write me own Science Fiction book. On the face of things R42 appears to be focused primarily on picking a genre that is going to sell at a particular time acting as a prompt generator for the author to shove up the chosen LLM.
For me that is counter intuitive to creativity but let's run with the profit motive.
Then things go downhill where it is suggested that the LLM, for cost purposes and because they are useless at maintaining focus is going to write or propose 1,000 word chapters. For me ChatGPT 4 completely barfed all over the plot I gave it. No intuition. No creativity. Just a mish mash of generic stuff it had previously been fed. Horrible.
And for non science fiction as focused on in the presentation? Well. Good luck dealing with the hallucinations. Your best seller on Care for The Aged might generate a few law suits. It's an accident waiting to happen.
Sign In
Follow the stocks
