Oh, wait! I don’t need to understand advanced mathematics to enjoy deep machine learning. I don’t even need a massive dataset. All I needed was three quick tutorials to realize.
I had my job taken by robots, and it feels okay
I had my job taken by robots. Like, seriously. I was a reporter at a newspaper up-north in Sweden, writing game-summaries of local low-division games to be found at the back of the sports-section, right under the results and the following standings. Oh, how the 14-year-old me loved that job!
All was done manually: At ding-o’clock, I rang coaches of both teams and asked about the game. I rang the referee to get the score. I synthesized the information to tell a short story about the flow of the game, the heroes and the atmosphere around the field. The emotion of teams fighting hard, to win and advance through the divisions or escape degradation.
Fast-forward less than a century: the same read supplied by a program, scraping the statistics and automatically writing short sentences about the outcome.
Unlock potential with the platform
That’s why I started using Peltarions platform. To stay ahead of the economic curve – to make my own programs, that achieve awesome stuff. That’s why I wanna talk about what’s to find between the lines in the first tutorials in the Knowledge Center.
In the beginning, after making a program that recognizes numbers in pictures and another that knows whether a review is expressing positive or negative emotion – I did not understand how much I had learned. What I was capable of creating, already.
The penny dropped just after my third tutorial, which instructs you to iterate the model; setting your experiment off for another round of fine-tuning.
The eureka moment
I realized two things:
x1 I don’t need to know probability theory or conquer any advanced mathematical skills to create serious digital tools that solve real-world problems for me. I just need to get familiar with the components in the Modeling view and experiment my way to evermore successful outcome.
x2 I don’t need an extensive dataset. Peltarion provides state-of-the-art-models that are pre-trained and ready for me to use a small dataset to make my very own specialized model. Furthermore, there is an extensive dataset-library already provided on the platform, with clear instructions about licensing.
Now I dream about taking back the magic to my technological world (which for some reason so far just has enhanced my world with ads, noisy information and social weirdification). My very own AI-driven image-identifier of what trees are around me. That’s what I’ve been setting off to learn myself for summers on end but haven’t managed. An application founded on advanced, powerful Deep Machine Learning code. To make the world around me more magic, and richer.
I know a thousand brands but I don’t know the name of the trees populating my home-towns nature. The bushes, the flowers, you know, everything that Carl Linnaeus named (by the way, here is a taxonomy for the third millennia). Maybe I can make the AI tell me their name just by showing them a picture, and all of a sudden, I would walk around in an enriched world where a tree is no longer just a tree among others.
02/ More on Business & AI
Get started with AI-kit
5 reasons you should take the leap from Plus to Pro this fall
Peltarion No Code AI Accelerator Program 3rd Edition
Peltarion No-Code AI Accelerator Program
Peltarion’s response to the Swedish Ministry of Infrastructure’s consultation on the EU’s proposed artificial intelligence act