Advertisements

Tag Archives: Machine Learning

RAIDI

Robot and Human hands touching

I have no doubt I am a very lucky person. Although I do not have an education in any science, I was able to spend approximately two decades working on the Space Shuttle Main Engine (SSME) program at Rocketdyne (through four major aerospace corporations). I spent a lot of time working with some of the brightest rocket scientists (for realz) as well as world-class engineers and scientists in literally dozens of disciplines.

Since my retirement from (what was then) Pratt & Whitney Rocketdyne, I have worked intermittently with Quantellia, LLC, an artificial intelligence / machine learning software development firm. Needles to say, I have no formal education in any computer field, with the exception of two Visual Basic classes I took at a nearby Junior College. I was introduced to one of the co-founders of Quantellia shortly after my retirement. She showed me a tool they had been developing called “World Modeler”. It was the most exciting thing I’d seen in a long time, and I was especially impressed with how it brought a highly systemic approach to modeling and forecasting in complex situations. I ended up writing several papers and a bunch of case studies for them.

In 2015 I returned to work at what was then Aerojet Rocketdyne (still is, for now) where I worked on a small rocket engine program for a little over two years. After leaving, I started doing some selling for Quantellia and, beginning in March of 2018, I became the company’s Business Manager, a position I’m still working at.

Last year we held a summit, in conjunction with SAP Global Services, at their Labs in Palo Alto. It was called the “Responsible AI/DI Summit.” In this context AI stands for “Artificial Intelligence” and DI stands for “Decision Intelligence.” One of the main purposes of the summit was to discuss how we can develop artificial and decision intelligence such that we concentrate on using them to solve humanity’s most “wicked” problems, rather than merely work at developing apps, the main purpose of which is to make money for the developers, investors, and entrepreneurs involved in the business.

Below are some of the folks who worked on the Summit, including me (the long-haired guy in the middle of the back row). Also, here’s a link to this year’s second Summit – Responsible AI/DI Summit 2019, as well as a link to the RAIDI Blog.

Quantellia and SAP folks who worked on putting it all together

As I learn more about machine learning, artificial intelligence, and decision intelligence, I will work at sharing my knowledge and understanding of these tools, and the issues they raise. I know the people I’m working with are dedicated to serving humanity, not merely milking it for profit. That pleases me and I hope we’ll be able to prove we’re doing the right things to ensure such service continues to exist and grow.

Advertisements

Will Someone Stand Up?

In the 2020 General Election, coming up waaaaay sooner than you think, time being what it is, there are eight (count ’em, eight) Republican Senators who are up for election unopposed. Actually, two of the eight are retiring but, in all cases, whether it’s a replacement or the incumbent, they’re all running unopposed. This is an intolerable situation, IMO.

Allowing any Republican, all (save for Justin Amash) of whom have shown themselves to be hapless sycophants, bowing to the whims of the most destructive and inhumane President in modern history, to run without any Democratic opposition is something we should avoid at all costs.

  • Bill Cassidy, Louisiana (In 2014 he beat three-term incumbent, Democrat Mary Landrieu, 56 percent to 44 percent. Don’t know if there are any Democrats in the running at present.)
  • Mike Enzi, Wyoming (Retiring – This seat is considered safe by most people.)
  • Cindy Hyde-Smith, Mississippi (Hyde-Smith defeated Mike Espy last November in a racially charged campaign.)
  • James Inhofe, Oklahoma (This is the schmuck who brought a snowball into the Senate chambers to make the argument that global warming can’t be possible because it’s still cold somewhere.)
  • Pat Roberts, Kansas (Retiring – Maybe a lost cause, as he ran unopposed last time and Kansas is a deep red state)
  • Mike Rounds, South Dakota (The entire state has approximately a quarter of a million voters. Unknown if there are enough Democrats to matter.)
  • Ben Sasse, Nebraska (In the 2014 election, there were a little over a half million voters; Sasse won every county in the State – 64% to 31%)
  • Dan Sullivan, Alaska (In the 2014 election, Sullivan won by 2.2% with a total of only a little over a quarter million voters. This state could be ripe for a flip.)

After the 2016 General Election, I worked with a group of people who were creating a canvassing tool that was designed to use AI to better prepare people who were out knocking on doors. It would have used demographics and historical voting data to train a machine learning algorithm on the patterns to be found in the data. Unfortunately, our primary investor kept adding requirements and ultimately squeezed the value right out of the app.

Nevertheless, our original concept we had discussed was to use machine learning to help political organizations make the most effective (not merely efficient) use of their various resources, e.g. time, money, people, connections, as well as understanding the political environment based on polls and overall news coverage.

Frankly, nobody I know of has sat down and begun to develop such a decision model, though I would dearly love to see it happen. It’s what we envisioned after Trump “won” and I still think it’s a viable approach. It does look like it’s a somewhat daunting challenge, however, when it comes to how expensive it would be to gather all the data we’d need access to, as well as develop the algorithms that would analyze and correlate the data.

Regardless, it seems a shame so many Republicans might run without any Democratic opposition. You’d think the least we can do is make them fight for their seats, which would include forcing them to shift resources around as well. It should be part of the overall pattern of the elections, which I’m unconvinced the Democratic Party really understands.


Hey! Long Time, No See.

QuantelliaLogoPaleI know it’s been quite a while since last I posted here. I’ve been continuously active on Facebook and have begun tweeting quite a bit as well, but that’s not why I haven’t posted to this blog in the past nearly three months. As of March 1 I began a new career, probably not the kind of thing you hear about 70-year-olds doing all that often. Since then I have been working as the Business Manager for Quantellia, LLC. You may recall I’ve done work for and with Quantellia on and off for the past six years.

Quantellia is a small AI/ML software development house and, until now, one of the co-founders has been running the business. Inasmuch as she is also the organization’s Chief Scientist, and a well-known pioneer in Machine Learning, this was not exactly the optimal thing for her to be doing. I had been touching on the subject and, since she was having such a hard time getting someone competent to run the business, I pressed my offer to do so. She finally relented and things have been going swimmingly, although there have been times I was swimming against the current. I’m definitely climbing a steep learning curve, which sometimes has me questioning if I’m losing my edge.

Actually, at times I can’t quite tell if my intellect is slipping a little bit, or if I just don’t care as much as I used to and I’m not quite as arrogantly sure of myself. My memory seems to be intact, along with my ability to learn and adapt. I’m going to go with the “I just don’t care as much about things as I used to; I’m more sanguine about life, work, and the need to control everything.

At any rate, I’m having a lot of fun. I was once partnered with two CPAs, doing royalty accounting for some big acts: Jackson Browne, Joni Mitchell, The Cars, Dollie Parton, Ronnie Milsap, The Commodores, even Jimi Hendrix’s estate. I learned a fair amount about accounting back then, and now I’m getting the opportunity to revisit what I learned, applying it in different circumstances. I’m also learning about artificial intelligence and machine learning, and hope to convey some of what’s going on in these fields. Although not a data scientist, I am quite capable of seeing where AI can be applied in business to assist with all kinds of issues. I’m sure you can as well.


It’s All Over, Humans!

Despite this foosball playing robot’s ability, it will be a long time (if ever) before robots or artificial intelligence actually displace us. More likely, they will augment our capabilities and free us up from most of the drudgery we’ve been dealing with for millennia. I find it easy to imagine a day when humans will evolve into cyborgs; part flesh, part machine. In some ways it’s already happening with prosthetics and it’s almost certainly going to accelerate with DNA testing and in utero surgery.

Stay intelligent, my friends. 


%d bloggers like this: