Wednesday, December 8, 2021

Microsoft and the AI Approach To Creating A Better Company (And World)

At Microsoft Ignite this year, Mitra Azizirad, Microsoft’s Corporate VP of AI and Innovation, gave one of the most exciting presentations. She opened with a story about how the man vs. machine chess match was eventually won by a computer, one that became unbeatable until they figured out how to partner a top player with a computer that won. This win showcased that, together, AIs and experts are better than either is alone.

This blending man and machine concept has increasingly been driving development because it ensures future employment and ensures more effective solutions if both elements are well trained and adequately integrated.

Let’s talk about some of the things that this blend of man and machine will eventually create.

Instant Stories

Instant Stories is where an author would come up with an outline for a story, and the AI would write the book or script, completing the effort.

I expect you’d also put in parameters like how long you’d want the book or script to be, how racy, how much violence, and who the target audience was before letting the AI do its thing. Can you imagine creating a marketable book or script in a few minutes more than the time it took to flesh out an outline? Granted, you’d still want to tweak it.

Still, the AI could use its knowledge of what constituted a bestseller book or a hit movie to assure a more successful outcome far better than a human could. The creativity of the human could do a far better job with the initial concept.

One other aspect of this is the AI could create alternative endings for stories and, if tied to computerized image generation and Deep Fake-like capability, could even re-render a movie with that ending. This combination of technologies suggests a future where an increasing number of movies you watch at home is uniquely customized to what you would most enjoy. You could even have a situation where another director could near-instantly change a movie that tests poorly to be more successful.

Auto Email Response

Another use for this blended AI human capability would be to email where the system could, after reading an email and knowing a great deal about you, draft a response that you’d edit and send. This capability would make even long-form email responses possible on a Smartphone and allow you to better interface with the device for an email with voice by speaking to it – much like you would have a secretary back when we had those.

Super Digital Assistant

If you are like me, you often ask your Digital assistant questions about the movies or TV shows that you are watching and finding that it isn’t that good at finding the answer (or even understanding the question).

Mitra showcased that Microsoft’s AI not only could answer those questions, but it could also find parts of the movie or show you were asking about and use that segment as part of its answer. It even knew not to answer questions without an additional prompt that might result in spoilers.

For instance, if you asked where Tony Stark had serious conversations in Avenger’s Endgame, it could find every instance where that occurred. It was able to answer a general question like how many people did Thanos kill in the movie. Now imagine that same capability applied to taped meetings or even documents under review for litigation discovery.

The concept of an AI that not only could accurately transcribe voice to text but understand the nuanced meaning of the conversation to identify what you were looking for is game-changing.

Project Origin

Project Origin has been created to fight the proliferation of Fake News and Deep Fakes, which is tearing many countries apart, including our own.

This effort is designed to preserve news and videos’ integrity to trust better the information we are given. It prevents manipulation of media between the creator of that media and the people who consume it. This data integrity solution would prevent many of the efforts to reframe or change a news story to change the event’s perceived outcome and perception. We were using a process that is surprisingly similar to Blockchain with distributed ledgers. It makes altering that source information without being caught almost impossible.

Human AI Training

During the event, we were able to train a virtual drone. This example was to showcase how people could train autonomous vehicles by just taking them over and showing them how to behave.

The nice thing with AIs is that once you’ve trained one, you can almost instantly train every other AI doing the same job. If the autonomous device is destroyed, you can replace it and its training rapidly using the earlier training information.

Generate Learning And Testing For Each Employee On Demand

We are all different, yet our education programs are generally static and inflexible, leading to training and education failures.

Imagine a system that could customize at a granular level a training or even a complete education program that would best match your unique needs. This program could also alert you, before you started, that this isn’t an ideal path for you and suggest another course of study or a different type of job.

Homomorphic Encryption

One of the enormous problems we had finding remedies to Covid-19 was that the massive repositories of medical information are encrypted and surrounded by policies and regulations that prevent global analysis.

Homomorphic Encryption, which you’ll be hearing a lot about going forward, allows analyzing this data where it resides without decrypting it. All of the identifying data tied to a patent remains protected and unseen. Simultaneously, the analytical effort looks for trends across patents that could result in locating a drug that could mitigate or cure a disease sent to the researcher. This technology could be critical to mitigating the next pandemic.

Wrapping Up

Microsoft showcased a robust future at Ignite this year, and in no presentation was that more evident than in Mitra Azizirad’s talk on blended AI. She showcased how applied AI could create more customized experiences in training and entertainment, create a super Digital Assistant that could answer even the most obscure questions, how we could better fight Fake News, and how we could more effectively fight the next pandemic.

She spoke of a future where instead of being afraid of AIs, they became our partners in our fight to create better, happier lives, more productive companies, more honesty and truth in the news, and a far better tomorrow. Suddenly, Microsoft looks like it is the place to be if you want to change the world.

Similar articles

Latest Articles