Ann: Appendix 3B Convertible Note Conversion, page-8

Hi homeales.

I appreciate the point your making, but in all fairness it's probably going to take a bit longer than 6-10 weeks to know whether we're on a winner with this. Sure we'll have proof that the thing works in silicon. And I'm 100% confident it will.

But we really wont know whether we've past any winning posts until customers have had their time to put the engineering samples through their paces. On the plus side, when this stage is completed it's likely Peter's Akida chip will be viewed like "unicorn technology" by those in the know in AI. (It will be a pity that the rest of the world wont have a damn clue what it's all about.) Anyway, because of the disruptive nature of ditching deep learning, for something new, it's likely our sample chips will face more rigorous testing than might be expected, and then extra verification and validation before anyone spends a buck on any custom design and integration. In view of all this I think we all appreciate now this will not happen overnight. And in this respect, assume much of the same re the time involved putting together our first licencing deal involving Akida's Intellectual Property.

Cheers
Cyber