[ad_1]
Image: © AFP
The Washington Post innocently published an article called “Is it A.I.?” The idea was to find out what experts and WP readers thought was A.I. in specific applications and what wasn’t. Sometimes they agreed, sometimes they were way apart.
You could call it “experience lag” for users of anything. Most people use a variety of vintages of software. All software has its quirks. Gaming software in particular delivers a vibe where you can hardly avoid assuming that it’s doing its own reasoning.
At the moment, A.I. as a class of technology is borderline. It’s integrated into some things but not others. There’s a physical cutoff point with older software which simply doesn’t run like that and was never designed to run like that.
Functional operations like tap-to-pay credit cards were generally considered by A.I. experts and WP readers not to be A.I. There’s a good reason for that. You’d have to integrate A.I. of various flavors with the global financial system, SSL, and other critical functions. If you didn’t the result would be grim at least.
A.I. has to work with a vast range of functions, generalized and/or specialized. Integration is going to take a while for that reason. …But it will happen.
The most likely scenario is that the already mostly automated systems will integrate first. Interactions with users are relatively minimal. Pay a bill, use a chatbot, or order something on Amazon, it’s easy enough, and low risk to businesses. If there’s a problem, you can find it and fix it with minimal issues.
The finance sector, that famous global monastery for altruists, is in the firing line here. You’ll notice these simpler things require almost no user discretion. That doesn’t work in other areas of business. Stock markets, futures markets, commodities markets, credit markets, and other high-risk markets can’t work that way.
In stock markets, auto-selling is fully automated but discretionary. You can buy at X and sell at Y with no issues. Add user interactions with A.I., and you get a very different world.
Markets are behavior-driven. They’re excellent behavior studies in their own right, without any technology. The depths of human idiocy are always visible. Add A.I. and inexperienced users and inexperienced markets, and things can get messy.
Now add hype. As usual, unproven tech takes center stage based on nothing else. “Our new whizzbang handy dandy A.I. can make you a stock market billionaire overnight”. See any issues? Put it that way, anyone can see issues. Add a market expert and recommendations from a day trader, and you have the ideal scenario for absolute chaos.
This is where A.I. is dangerous by definition. The expectations of A.I. as it is now are ludicrous most of the time. It can analyze markets. It can do reporting. So can people who’ve been doing that for years, and they often as not misread market moves.
Currency markets are another hot topic for A.I.. A.I. is currently down valuing the Australian dollar based on a range of indexes. The AUD was already well down and has been for years without this type of very conventional analysis. A.I. doesn’t have “market sentiment” and unlike forex investors, isn’t necessarily looking at other investment options. Take half a second to figure out what sort of training is required for A.I. to become proficient. It won’t be overnight.
What goes in is what comes out. Training goes only so far and has to give way to experience. In finance, that could get expensive. A.I. may or may not factor in market news. It may or may not do projections based on anything but standard valuations. If A.I. were to analyze Truth Social, what would it do? Markets were short-selling it before it even listed. Maybe that’s a target for A.I. learning about markets.
In media, it could be pretty funny. An A.I.-generated movie breaks all box office records. So the markets might downgrade the stocks of media companies, based on outdated technology.
The good news is that there is plenty of time for being picky and choosey. Rushing in is more likely to be expensive than efficient. Most of the markets don’t actually need A.I. for daily operations. Data volumes and handling will more likely drive initial A.I. usage, rather than the bread-and-butter functions.
…Which brings us not-all-that-elegantly back to the question – What’s A.I. doing in your life right now? Probably not very much – yet. There’s no glaring need. There hasn’t been enough time to fit it into all the core functions of the society.
A.I. is currently an infant. It’s clumsy. It doesn’t know its way around basic things. It will grow up, but what’s the hurry? To deliver what it can deliver, quality of service and outcomes are far more important. Let’s see if anyone in the investment market notices.
_______________________________________________________
Disclaimer
The opinions expressed in this Op-Ed are those of the author. They do not purport to reflect the opinions or views of the Digital Journal or its members.
[ad_2]
Source link
