Appen (ASX: APX) (OTCPK: APPEF) shares are down 32% today APX shares fell on the results. Which show something that no one really wanted to see. Here we are in the middle of a boom in artificial intelligence, the company does artificial intelligence and yet revenues are down, earnings are down - negative in fact - and in general, well, what the heck’s going on here?
The results announcement: “Group revenue of $138.9 million decreased 24.0%, primarily reflecting a lower contribution from Global Services which recorded a revenue reduction of 27.4% to $100.1 million. New Markets revenue2 of $38.9 million decreased 13.7%, primarily impacted by a lower contribution from Global Product.Underlying EBITDA (before FX) of ($15.7) million compared to an Underlying EBITDA (before FX) of $9.6 million primarily due to reduced revenue and gross margin, and a proportionally higher cost base coming out of FY22.Underlying EBITDA (including the impact of FX losses) of ($18.1) million, compared to
$8.5 million. Underlying net loss after tax4 of ($34.2) million, compared to an underlying net loss of ($3.8) million. Statutory net loss after tax of ($43.3) million compared to statutory net loss after tax of ($9.4) million.”
Well, no, that’s not what anyone wanted to see at all. The difficulty is this: “Appen Limited, together with its subsidiaries, operates as an AI lifecycle company that provides data sourcing, data annotation, and model evaluation solutions. It operates through two segments, Global Services and New Markets. The company offers data sourcing services, including image, video, speech, sensor, multi-modal hardware device testing, and mobile location services; pre-labeled datasets of audio, image, video, and text; and language-based AI solutions” Everyone and their grandmother are rolling out AIs. Why can’t Appen make money at this point?
Appen share price from Google Finance
The underlying problem here seems to be that Appen provides the data sets and services which train AIs. But past a certain point of AI development it’s not, in fact, necessary to use carefully curated data sets and services to train an AI. Or, even if it is, too many people are running with not having to do that.
That’s one reading of it at least and that’s the one we’re running with. The other one to consider is that AI is now considered so important that people aren’t going outside the AI firm for those data sets and services.
Either way this does produce a problem for the long term of Appen. If it can’t make money right now in AI then when will it ever be able to do so?