Would you pay for AI courses / software ?

[Nosk]

Hello everyone,

I finished my PhD in neural networks (particularly auto-encoders) 2 years ago and have a steady and nice paying jobs.
However, I'm thinking about doing some extra work on the side to get a little bit more btc.

Would you consider paying for course, models/software, or just me thinking about a solution for a problem you want to solve ?
If yes, how much ?

I do have a PhD and several paper online (which I can explain if you want) but I'd rather keep them secret, so as my identity.

EDIT : I'm familiar with neural networks for classification, regression and anomaly detection. I also practiced various machine learning "simpler" models and ways to pre-process data.

[SFR10]

Would you consider paying for course, models/software, or just me thinking about a solution for a problem you want to solve ?
If yes, how much ?

Well that answer it 

Don't be discouraged by the lack of responses so far ^{[this topic has been viewed only 19 times so far]} and just to give you a general answer, you'll have a real chance in charging someone for a course ^{[and etc...]} if a similar one doesn't exist already on the internet for free or for a cheaper price than yours. In regards to your last question, that depends on the complexity of the problem ^{[don't expect others to answer it accurately]}.

[Nosk]

Well it is easy to find cheap ML engineers online but what they do is mostly
 

• use guidelines to select a model (google "machine learning cheatsheet")
• pre-process data in an automatic (/"dumb" ?) way
• feed the data to the model, then go "tada, you can have 80% acuracy for your problem".

I have several advantages : I have experience in various domains (Image or text classification, time series treatment, anomaly detection, image clustering)
+ this is my main job so others are already willing to pay for it
+ I can EXPLAIN what I do, why I do it and what the limits are. If you just want a small course on basics, I can do it (I have already given courses to students of various levels during my PhD)

Example : Your problem is
"I want to predict the price of the bitcoin"

1st thought : Machine learning aims to explain / predict / estimate results based on data you have. Bitcoin price (and all stock market-like time series) are explained by themselves AND external factors (you can't predict china ban / elon musk fud without looking at very different data).
So don't expect a very accurate result if you want to stay full chartist.

2nd thought : When possible, simplify your problem. Predicting the price may be optimistic but predicting the direction may be sufficient for your goal (i.e. become filthy rich with your magic box predicting bitcoin price). May be you can start here, then try to predict what the volume will be, then maybe try to crawl social network to pre-shot external factors, etc...

3rd thought : Having results based on a initial confidence ("when I trained my model I had 85% acuracy") isn't reliable in time. You must think about a way to have a confidence factor during the usage of the model. Example : think about 2 models (similar or not) predict more less the same thing (can happen a lot in machine learning). When 2 monthes passes, the models start to disagree. It means that you can't rely on the results of your models anymore, and you need to do some fine-tuning / re-training / ...



Price is debatable, I know that my company "sells" me for 1k€ / day to clients, freelances sell themselves for 600€/day, but everybody take much less because of charges + company takes money (I gain ~200€/day). + you can find cheap and dumb AI engineers online, as I mentioned before.

Just don't go "I want a full neural network translating english to japanese in real time with 99.9% acuracy, trained by a GPU rented by you, for 300€" please 

[Nosk]

If some hesitate, just PM me and ask "Hey, I have this and I want this, is it possible ?" or "What do I need to do this ?" or "Can you do this and that ?"

[vv181]

You explain thoughtfully the disadvantages of others within your domain. And the only one who can fairly set the price to monetize your knowledge/work is yourself. I believe if you are able to gain trust and respectable experience for your work or portfolios, you will able to monetize it in easily manner, not to mention that you already comprehend what others lack off.

I know that machine learning has a promising future, and maybe if you collect some portfolios within the Bitcoin niche for your skill, it will get you a respectable trust, and others will probably believe you for your skill. Try to make an announcement or service offer in the appropriate section. Let the people know that you have the potency within your domain.

[maisao1991]

What are the courses meant to provide and is it chargeable? there are a lot of AI courses for participants that are very useful and we want to talk about it. can you bring this to your course?

[Nosk]

I can do courses about various subjects.

By level :

• Beginner - I'm in IT since 11 year and in Machine learning since 6 years, so any subject you bring to me and let me prepare, from algorithmics to databases or genetic algorithms (just don't ask me stuff about websites, especially front). This level should be cheaper.
• Medium - I can cover machine learning basics and generalities (you'll know what is "classification", "regression", "dimension reduction"... how you setup a machine learning project, and more) to specific model (I practiced SVM, random forests, various interpolation methods, PCA, etc etc). In this level, I can think about the skeleton of the project you need and I can explain the math or simply what are the specifications of each model you may want to use.
• Advanced - I can explain in a very detailed way how neural networks work, what kind of NN exist and when you should use them.

[GreatArkansas]

The courses are good but you have a lot of ways.
Creating a software or Software as a Service (SASS), which could lead you a passive income once you will get users but I know having a SASS is difficult and complicated.
Another thing is tutorials, although it's kinda a course too, but what if you do tutorials based on 1 on 1, or by group live?

[fearfulcalm]

If it have a good content, probably yes
but if it's junk stuff i prefer to research myself at internet

[Naficopa]

I have interest in AI as its coming like a hurricane. Do you have any outline of A.I courses you are offering? I am a newbie so will be interested in learning A.I from scratch. Hope you have relevant material for that. Maybe you should think about some demo material.

[semobo]

Yes, you can make extra money by teaching the things you know on an online platform which is already happening in all the platforms and YouTube is one of the main source for it, you can teach things free of cost to the viewers but still you can make money from the ads revenue or you can just teach them the basic and ask them to visit yoir website for complete course amd you can set the subscription fee for that.

[Nosk]

I have interest in AI as its coming like a hurricane. Do you have any outline of A.I courses you are offering? I am a newbie so will be interested in learning A.I from scratch. Hope you have relevant material for that. Maybe you should think about some demo material.

I haven't done any material yet, but A.I. introduction for newbies surely is a thing I can do. You're right, maybe I should do a portfolio/demo website or something like that.

Yes, you can make extra money by teaching the things you know on an online platform which is already happening in all the platforms and YouTube is one of the main source for it, you can teach things free of cost to the viewers but still you can make money from the ads revenue or you can just teach them the basic and ask them to visit yoir website for complete course amd you can set the subscription fee for that.

Well doing videos on youtube and expecting money from it sounds quite terrible (plus my purpose is to get paid in btc and not to have to bother with taxes and such). Not bad for demo material though.


We're all talking about "material" here but at first I was more interested about giving like few hours on discord of "real" class + presentation / hans-on coding. I can also probably just "sell" formation, but there are already some great ones online (in the non-trainer area). Or maybe this can also be part of the demo material.

What I take away is that it seems that people here are more interested in learning than in realizing or getting help carrying a project.

[hugeblack]

Since this project will not be your main job, do not seek to make a return from it, but present the idea first, find solutions and the returns will come on their own.
I do not advise you to enter the field of forecasting, especially the price.
You can create a program using artificial intelligence so that it is free and open source, if there is a need for physical parts, manufacture them at a low price and provide some paid features later if you get a good downloads.

[dbc23]

AI courses are hot cakes and lucrative. I would love to get PDFs if you have any available or a recommendation on where to offer such courses either for free or for a reasonable and affordable token

[ali1177]

I am artificial intelligence enthusiast if you will provide good content through your course then definitely I will pay for the course but first I have to check the quality of the course.

[Ekeleboy]

Why not.
AI seems to be growing fast and covering grounds.
Anyone wouldn't want to miss out.