Weve been overselling running capabilities of AI for years but that doesnt mean it doesnt have a shining forthcoming. Thats possibly why Stanford University researchers conceived of a ’One Hundred Year Study on Artificial Intelligence’ (100 years!) back in 2016 with plans to update the report see five years through 2116 charting the progress of AI along the way. Five years behind the inaugural report the study authors recently released the second report.
The TL;DR? Weve made ’observable progress’ in just five years on the back of ever-improving data infrastructure yet we quiet fall ’far brief of the fields founding longing of recreating full ethnical-like intelligence in machines.’ What we are discovering however is the weight of meshing ethnical and machine to accomplish better outcomes. Is it ’true’ AI? Not as originally envisioned. But arguably its better.
One of the leading inhibitors to data science (and resultant AI) beseeming real has pliant to do with science and seething to do with data. As FirstMark investor Matt Turck recently named out in ’The 2021 Machine Learning AI and Data (MAD) Landscape’ only recently have data warehouses evolved ’to store solid amounts of data in a way thats advantageous not fully cost-prohibitive and doesnt demand an army of very technical nation to maintain.’ Yes weve had data warehouses for decades but theyve been confused and precious. More recently weve dabbled in Apache Hadoop which made things cheaper but quiet overly intricate.
Only in the past few years has the activity centreed on maturing our data infrastructure such that it has befit dramatically more approachable for mere mortals (who may or may not have a PhD). By making it ’finally practicable to store and process big data’ in a cost-effective mode Turck argues it ’has tryn to be a major unlock for the rest of the data/AI space’ in three leading ways:
Although Turck chooses to centre on the real contact of present data warehouses the activity has also benefited from other advances in databases (distributed databases NoSQL etc.) and the cloud which has made it easier to iterate on data. Through these and other forces it has befit easier to store and work with data which in turn has enabled organizations to do more with that data.
Which brings us back to Stanfords AI100.
Weve reached a point where we interact with AI on a day-to-day basis and commonly see its briefcomings. Take Tesla. For all its mismarketing of AI-infused ’full self-driving’ Tesla electric cars are nowhere near being capable of safely taking passengers from point A to point B in anything but the most carefully controlled environments. Even so weve seen sufficient to be intrigued and hopeful for the forthcoming.
In the present the AI100 authors point to three areas where AI has demonstrated real progress:
This doesnt mean AI will restore ethnicals anytime soon but it does mean that AI is increasingly capable of complementing nation in meaningful ways. As they expound ’AI approaches that augment ethnical capabilities can be very precious in situations where ethnicals and AI have complementary strengths. An AI method might be better at synthesizing useful data and making decisions in well-characterized parts of a problem while a ethnical may be better at knowledge the implications of the data.’
For sample the reports authors say that machines will never be a suitable exchange for nation caring for the seniorly. ’Good care demands respect and dignity things that we simply do not know how to code into procedural algorithms.’ But AI that crunches big quantities of data to hint to caregivers when an senior may need remedy or other support? Or possibly using AI-driven image processing to evaluate medications the senior may be taking on her own but that could try harmful (owing of measure or the essence of the medication itself) and alerting a caregiver? Thats a big union.
Sometimes the artifice is to set the AI standard free to analyze data then aspect out how it reached a conclusion. ’By leading training a standard to be very good at making predictions and then working to apprehend why those predictions are so good we have deepened our philosophical knowledge of seething from disease to earthquake dynamics’ the authors note. In this sample the machines push nation to ponder more deeply almost data learning from conclusions the machines dont apprehend but are able to arrive at anyway.
Machines in brief are able to analyze huge quantities of information summarizing or otherwise presenting that information to nation in a way that makes it more digestible. In this way ethnical intelligence can more effectively be applied. Humans wont restore machines and machines wont restore ethnicals. We build the data infrastructure that makes copious quantities of data practicable and the machines do their part by helping us to make perception of it all. A nice union truly.