The most immediate concern for me after the conference was to build a comprehensive summary of precepts shared on Scale AI – transform X conference; at least, nuggets pronounced by the infamous & thought-provoking speakers… then I realized that would be redundant. One of the rules of software engineering is to reduce redundancy. I won’t delve into being repetitive – and rather beseech you to go forth and watch the on-demand videos.

Next I had to come up with an architecture to break down my commentary of the entire event; which stood out for me quite superbly – from onset to its closure. One strong reason to participate in such industry wide occasions is to gauge the usage of AI technologies. AI being the core theme of Scale AI – transform X – conference.

Which brings me to key use cases presently across various disciplines as showcased at the event – from AV(autonomous vehicles), to health care, to finance, to online commerce, to national security, to research, to data synthesis and lastly to enterprise implementations. AI tooling is what each industry sector will have to deal with sooner rather than later. For whence data dwells – it remains untapped if not subjected to Machine Learning/Data Science algorithms for insight(s) generation. In essence, unveiling the secrets of the data, if any.

After that, I’ll embark to highlight some of my outstanding takeaways from the conference. Such as the ‘twining’ terminology usage. I found that rather pleasing. The whole concept of running simulations, based off of real life events to probe for more predictions from a limited dataset. Also the panel that talked of AI biases, was remarkable. It was one if not the most passionate session. The speakers emphasized some dire recommendations going on forward for AI usage.

One of the downsides which is subjective to my experience: was the global outlook by Eric Schmidt. I found the whole sedimentation remark: to nations clustering behind either the USA or China to be quite limiting. In a way my intuition repelled the whole concept. AI ought to be the great unifier – call this notion naïve – but, I don’t see how rallying predictions within this antiquated approach benefiting humanity as a whole. And anyone who is progressive enough would know some of the biases spoken at the conference get reinforced because of propagating such misplaced metrics! Which shouldn’t have been center stage at the conference, but that’s just me hey! I like (and I’m sure I am not the only one) who loves what folks like Andrew Ng are doing globally for the AI community.

I will end this short essay with my favorite use case of AI thus far across our sample at the Scale AI conference. That will have to be Spotify implementation of recurrent neural nets, to embeddings(similar artists’ recommendation), to reinforcement learning(events across the platform) and other ML algorithms. The vignette was precise, concise and easily relatable by any attendant – regardless of their background in AI. I will hopefully react to other individual sessions from the conference in future to do in-depth analysis. Until then, keep pursuing your worthwhile AI passions!

Leave a Reply

Your email address will not be published.