A List of 15 Free AI Software Programs to Download
Table of contents
So you just received your MBA and now you’re thinking about starting an artificial intelligence (AI) software startup. Well, so is just about everyone else. If you do decide to build your own AI software startup, you will be joining the likes of +1,500 other artificial intelligence startups out there that have already thrown their hats into the ring. You’re going to need to move fast, and your future baby unicorn is going to require three basic things:
- Talent
- Hardware
- Software
The first one you’ve all but nailed. You’re an MBA graduate. What don’t you know how to do? At the very least, you should understand something about managing people now. After all those classes on leadership, you’re oozing with the charisma it takes to attract the best talent on offer. The second one is easy too. You can just buy some hardware from the market leader in AI hardware – Nvidia. The third one is really the meat of the problem you’re facing. You need to decide which machine learning software you should use.
These days you’re not going to reinvent the wheel but rather use an existing platform to develop your solution on. Most likely you’ll want to use an “open source” platform that you can download now and get the ball rolling. What open source means is that the software is free to use and you’ll be joining an entire community of people who are developing on the platform just like you. They’ll likely be a whole lot smarter than you, but don’t worry. You have an MBA now.
If you’re in a hurry to get your unicorn off the ground and you don’t have the time to read the rest of this article, here are all the open source AI tools we’re going to talk about along with our own insightful commentary provided by one of our on-staff MBAs:
| Name | Company | Insightful comments from an MBA |
| Tensorflow | Google’s ability to promote their machine learning framework and fund it ensures that adoption is going nowhere but up. Nobody has ever been fired for choosing Tensorflow. | |
| Caffe | Not-for-profit | Supported by grants from Nvidia and Amazon. Caffe can process over 60 million images per day with a single NVIDIA GPU. |
| Torch | Not-for-profit | Used within Facebook, Google, Twitter. Computing framework for machine learning and computer vision. Originally developed by academics at NYU. |
| Theano | Not-for-profit | A Python library that allows you to evaluate mathematical expressions involving multi-dimensional arrays. Whatever that means. |
| scikit | Not-for-profit | Machine learning in Python. If you want to see lots of cool examples of what can be done with scikit, just click here. |
| H2O | H2O.ai | World’s leading deep learning platform used by 80,000 data scientists in 9,000 organizations. Used by 20% of Fortune 500 companies. $33.6 million in funding. |
| Neon | Nervana Systems | The “world’s fastest deep learning framework”. Acquired by Intel in August of last year for $350 million. They’ll up-sell you to use their cloud once you build your app on their platform. |
| Deeplearning4j | Skymind | Startup which took a $3 million seed in Fall of 2016. Tencent lead investors. Deep learning library for Java that wants to be the “Red Hat of deep learning”. |
| CNTK | Microsoft | Microsoft’s “cognitive toolkit” that trains and evaluates deep learning algorithms with language like C++ and Python. Fast and scalable. Optimized for their cloud platform Azure. |
| Mahout | Apache Software | Quickly create scalable machine learning applications. Core algorithms for clustering, collaborative filtering, and classification. Improved sales from Amazon recommendation engine by 35%. |
| MLlib | Apache Software | MLlib is Apache Spark’s scalable machine learning library. The tagline reads “lightening-fast cluster computing” and speed seems to be one of their key selling points. |
| NuPIC | Numenta | CB Insights AI 100. Silicon Valley startup. Framework inspired by neuroscience called “Hierarchical Temporal Memory” or HTM. Invested in Cortical.io. Biologically inspired machine intelligence. |
| OpenNN | Artelnics | An open source class library written in C++ programming language which implements neural networks. Developed by a Spanish startup called Artelnics. |
| PredictionIO | Salesforce | Described as “MySQL for machine learning”. Open source machine learning platform for building predictive applications as the name implies. |
| Seldon | Seldon | London startup founded by data scientists. Full-stack open-source machine learning solution for recommendation engines. Serves billions of predictions a month. |
If you want to download any one of these, we’ve made it very easy for you by linking the company name to the download page. Which one should you choose? We wrote about Tensorflow before and gave all the reasons why we think it’s the number one platform you should consider. H2O seems to be getting pretty wide-scale adoption as well.
Conclusion
So what’s the takeaway for investors? If you were looking for a picks-and-shovels play by investing in the dominant AI framework that everyone will end up using, it seems like we’re a ways off from that. Is this a similar situation to when we had 10 different search engines before Google just dominated everyone? Maybe, and that’s probably what Google has in mind with Tensorflow. Note that Google is also adopting their own AI hardware as well. Of course, buying shares of Google in hopes of getting exposure to Tensorflow doesn’t make sense according to the “invest in everything with Google” fallacy. You should own Google for lots of other reasons.
While there may certainly be other deep learning frameworks out there, this list is not meant to be exhaustive. We simply wanted to give you a list of AI software tools and frameworks that you can feel confident running with and cost you absolutely nothing to download and start playing with. Any one of these open source AI frameworks can be suitable in helping you get your startup off the ground as quick as possible.
Share
hey you need to get your research done more accurately as the most impotent company you left out its more advanced then anyone of your ones that you have listed ……..just a tip check out BrainChip ASX code BRN you might find it interesting for your next article one day
More advanced than Google eh?
We’ll check it out. Thank you for the heads up!
I am looking any AI software that could be used in a gambling environment such as roulette.
A man after my own heart. This is an interesting question. If you had augmented reality glasses with a hidden camera that counted cards you could dominate the blackjack table. They’d probably be on to that so fast your head would spin. Wouldn’t roulette be purely random? Great thing to talk about in theory but obviously nobody should ever do something like that because you’re asking for trouble.
I have a panasonic touchbook CF31 blackhawk ( ex military ) fully kitted out complete with sat com
and 3G, I’d like to use it in conjunction with an AI for paramedic services. the idea that with its 3G
and satcom capability it could call for medical assisitence and give its geographical location
as wel as give firsthand medical advise for stablizing the inccident/accident victim …
Sounds sick. You can also get beacons that work pretty well for that same purpose.
i will like to know more about A I,thanks
So would we.
Free to download but commercially exploited. How come that all those similar blog’s and web sites only advertise commercial vendors, some even commercial ones as open source, but open source is entirely not mentioned once ??? And Libraries aren’t self contained install able options for 99% of potentially users. We just want to analyze data not making master degree on how to install and use it.
Good point