Latest news about Bitcoin and all cryptocurrencies. Your daily crypto news habit.
Welcome to AI! Welcome to machine learning! Does it matter if you donât know the difference? Nope, because youâll start applied projects in them the same way.
What way is that? Perhaps surprisingly, not any of these:
- Get an AIÂ degree!
- Hire an AIÂ wizard!
- Pick an awesome algorithm!
- Dive into the data!
Itâs a trap!
Look familiar? These tend to be favorite starting points, but theyâre all traps. Many businesses fall for them and fail at machine learning, but not you. Youâll start right.
Image: source.
But first, why are these the favorites? Itâs a story of comfort zones.
Your comfort zone can lead you astray
If youâre the studious type, your instinct might be to take a course or sign up for a degree. Watch out, though. The classic AI courses out there are probably wrong for your needs. Iâd hate for you to wind up with the wrong degree!
If youâre a business leader, your instinct is to hire someone who sounds qualified. Great instinct! Except the person best qualified to start an applied AI project is not your garden variety AI PhD. Itâs⊠you! Whoops. Hire yourself first and read on to find out what youâre supposed to do before you bring your champion nerd on board.
Donât ask a team of PhDs to, âGo sprinkle machine learning over the top of the business so⊠good things happen.âEerily familiar?
If youâre an AI researcherââârecently hired to sprinkle some machine learning magic on top of the business, am I right?â youâll want to start where youâre most comfortable. With the algorithm, naturally. You just spent 10 years of your life studying how to design new AI methods, so why would the leader want you to start elsewhere? Letâs pick an algorithm⊠mmmm, neural networks are all the rage. Maybe we can figure out how to make them even cooler? Letâs create a new approach! Now, what data can we shove into our new-ral network? (Hereâs hoping weâll wind up with something we can sell to the leader to justify the past six months spent inventing things.)
Or maybe youâre a data scientist. (Also a classic first hire, since todayâs market thinks data scientists walk on water.) Perhaps you also have a PhD, but your Great Love isnât methods. Itâs data. Data data data! What data do we have? Letâs figure out what beautiful ingredients we can use!
Wait⊠use for what?
If youâre a data scientist or AI researcher and this sounds familiar, you just got handed a lemon by your leader. They let you down! Go on strike until theyâve done their part.
Leaders, figure out whoâs calling the shots. If itâs you, then letâs designate you The Decision-Maker for this project. Otherwise, delegate the position to someone else and ask them to read the rest of this while you play outside in the sunshine.
Start here
Okay, Decision-Maker. It took a while to track you down, but here you are. You understand the business and you have plenty of imagination, so youâre qualified for this. Glad someone forwarded you this letter! Letâs get you oriented with how to set a machine learning (or AI) project up for success.
The right first step is to focus on outputs and objectives.
Imagine that this ML/AI system is already operating perfectly. Ask yourself what you would like it to produce when it does the next task. Donât worry how it does it. Imagine that it works already and it is solving some need your business has. (Thatâs why you needed those qualifications. Someone fresh out of a PhD doesnât understand your business yet, so theyâre not qualified for this task.)
The problem with the approaches discussed previously is that the order of operations is all messed up. The right way to approach an applied project is to flip the algorithms-inputs-outputs order on its head, like so: think about outputs, then inputs, then algorithms!
Your order of operations might be a mess.
A kitchen analogy comes in handy here. If youâre running a restaurant (as opposed to an appliance factory or food science lab), why would you think about buyingâââor, worse, inventingâââa pizza oven before youâve even considered whether adding pizza to your menu makes sense? That sounds like the rookie mistake of someone who doesnât know what business theyâre in. Instead, start with what your customers want and what food quality youâre willing to settle for.
Define success!
Figuring out what success looks like can be nuanced. Which of these three is good behavior?
âAll of themâ? But surely you donât want your police dog chasing sheep! Or vice versa, for that matter. A better answer is that it depends on what the owner wants. Thatâs you! Diving into algorithms and data before figuring out what outputs would count as good or bad behavior is a bit like putting a puppy in a basement with food and water, then being surprised what comes out isnât good at being a police dog. You canât expect to just sprinkle machine learning on your business, leave it brewing, and get something useful.
It took plenty of planning to get Peach this good at policing. He even writes witness statements! (This fist-crayon masterpiece comes from officers frustrated by a barrage of requests for an account from PC Peach⊠despite their having explained that Peach is not a person.)Analytics might a better fit for you
Applied AI requires you to have to have a very clear vision of what your model needs to grow up into (and why), then you have to train it towards that. If you donât know what you want, head to the font of inspiration instead: analytics.
Spend some time figuring out what looks promising enough to pursue, then come back to machine learning when youâre ready.
Besides, analytics uses some of the same math, so you wouldnât be lying if you told your friends youâre using ML/AI algorithms (though youâre not building ML/AI systems). Many people who think they want ML/AI actually only need analytics. The latter is a great idea for all projects, while the former is good for only certain kinds. If youâre unsure, go for the sure thing.
Is this good behavior? I have an opinion and Iâd hazard a guess that Fido here has one too.Before you do anything else
Dogs arenât born knowing that you donât appreciate their couch-chewing inclinations. Itâs up to you to think about what youâre looking for in a pet so you can train them towards your ideal⊠before you find out what the stuffing is made of.
The right time to think about your goals is at the very beginning, while your project is still a puppy!
What goes for puppies goes for ML/AI systems. To figure out what success looks like, you donât need to understand how the puppyâs brain learns from sensory inputs. You donât need to think about how those sensory signals are stored and processed (yet). What do you need is to figure out that you want a sheep dog (and what that means to you). To do your job thoroughly, you also need enough imagination to picture what behaviors youâre aiming for and what youâre trying to avoid.
Additionally, it helps to do a quick intuitive reality check: verify that relevant data is within your reach and that you have the hardware muscle to process it. If youâre training a sheep dog, are you confident you can get hold of enough actual sheep to show it? Even if you have sheep, your puppyâs brain needs to be able to take in and use information about them. If your âpuppyâ is actually a fly larva, itâs not going to be able to do good things with sensory data about sheep. (It canât run in production either.) I donât need to tell you that youâll have a problem.
Whatâs obvious with dogs seems to elude many ML/AI teams Iâve seen. Some only ask what the dog is for when they retrieve it from the basement after a few years. Well, now you know.
The right step taken by the wrong people
Figuring out what problem ML/AI will solve for you is the first and most important step in your project, but unfortunately itâs quite often taken by the wrong people in an organization. While itâs supposed to fall squarely within the decision-makersâ remit, for some reason leaders try to avoid their duties by hiring a bunch of PhDs and sending them off to âGo sprinkle machine learning over the top of our business so⊠good things happen.â What could possibly go wrong?
It takes business savvy to properly think through what an Ml/AI system is supposed to do for you and why itâs worth building. Focus on this first, before getting anywhere near the nitty gritty, including figuring out whether or not the algorithm thatâll solve your problem is considered AI or ML (you deal with that much later). If you have no ML/AI training, tackling this first part before youâve hired a team or bought sci-fi kit might sound daunting, but Iâve got your back⊠my next article will be a step-by-step guide just for you. In the meantime, you start brainstorming here.
The first step in AI might surprise you was originally published in Hacker Noon on Medium, where people are continuing the conversation by highlighting and responding to this story.
Disclaimer
The views and opinions expressed in this article are solely those of the authors and do not reflect the views of Bitcoin Insider. Every investment and trading move involves risk - this is especially true for cryptocurrencies given their volatility. We strongly advise our readers to conduct their own research when making a decision.