7/ Some reasons: A) inability to implement dramatic process change, B) lack of usable data, C) lack of access to both ML, engineering talent
8
11
71
Conversation
8/ 1 more reason: businesses today largely have structured or text data, not images, voice. NLU is the least far along of those 3 AI domains
6
18
82
9/ Of course, companies can (and should) certainly set out to capture more voice/images and progress NLP research
3
1
19
10/ This of course doesn't apply to some industries that DO rely heavily on images (e.g. radiology). But today that's a smallish subset
1
20
11/ RPA is NOT AI. It's fragile, hard to deploy, rules-based process automation tech based on GUI-level integration. More on this later...
3
4
21
12/ Most execs have very little understanding of how to "apply AI" today
5
11
42
12/ This is unsurprising: they are caught between the tech industry's "magic cognitive AI does everything" hype marketing...
2
3
32
13/... and of course not having the technical understanding to identify labeled datasets/decisioning opportunities for supervised learning
3
1
37
14/ AI-enabled consumer products are making much, much faster progress so far, delivering new UX to consumers who are voice and photo-first
3
3
40
despite training costs decreasing bc of obvious reasons (cheaper compute, DL), the consumer products have still required massive $$
1
e.g., apparently building just the acoustic models for Alexa took $50M (never mind language models, etc.)
hopefully, new enterprise apps can just build off the pre-built services from the larger players (tho those products rarely look great)