Any AI tech here?

machinelearning

#1

Just curious about the AI technology. If you are an AI Tech company can you tell us about your company and its usage? Or If you are also AI enthusiast lets talk about some next level innovation using AI.

Would love have general discussion on this topic.

Go!


#2

A basic premise i stumbled upon a couple of weeks ago

I believe the same is true of AI as well.
As a background note It was while researching GPU programming on the NVIDIA website that i stumbled upon an contributing IIT Alumni’s profile at Cornell University, where they are doing groundbreaking work in several fields of Technology, including AI.
As i watched Kavita’s talk on Youtube, it became apparent to me that it is the groundbreaking research in the field of image processing at Cornell, described in the early 80s as the Cornell Box that led to the creation of some of the worlds most famous computer graphics rendering engines like Quake.
Since last week im a fly on the digital wall at Cornell :smiley: . They have a page on Machine Learning in general and AI in particular which i think is quite useful to the AI aficionado, with a dedication to Frank Rosenblatt, the father of AI on the Machine learning page.
On the Machine Learning site there are numerous personalities who do give informative talks on ML, AI, NLP . There is also some dated course info from 2013 as well.

Again, from the NVIDIA site,

what i learnt about the importance of ML and AI is that the ultimate goal of computer science itself is to come up with a realistic alternative for the human brain, and in that pursuit ML and AI are the cornerstone of that initiative.

So what kind of machines do you think lie at the heart of Machine Learning in general ?
What kind of capabilities are we talking about when it comes to building such machines ?


#3

Your extensive research shows how amusing it is to you too btw one of my friend just graduated from Cornell University.

For me it all started almost 2-3 weeks ago when my friend who works for Gupshup started blabbering about chatbots/bots how they are the next big thing and he even told about the deals they closed with huge conglomerates. That lead me to learn about about chatbots and experiment with couple of chat bots on Facebook messenger after experimenting came the understanding its mechanics and it lead me to AI, ML and NLP. I like TED Talks so i started looking for AI ted talks and stumbled upon Andre Leblanc explaining Artificial intelligence and its future - https://youtu.be/xH_B5xh42xc here he talks about how fast this technology will grow and future will be of AI machines that will invent and find cures for complex diseases without live experimenting on humans or animals from his talk i took away AI medical innovation, fast changing tech industry and the most fascinating was how our brains can be part of cloud system. Medical is one way to make use of this technology, I am very much mesmerised by Jibo - https://www.jibo.com/ - Its not like Google Home https://home.google.com/ or Amazon Echo https://youtu.be/KkOCeAtKHIc , But Jibo is different its a social robot that has loveable personality that you would want it like a dog with robotic skills that helps you get things done. With technology advancing day by day I belive even for a day if you won’t be able to cope with technology you would left far behind in this fast speaced tech world.

Google, Apple, Microsoft all the top Silicon valley unicorns are eyeing on AI Medical tech having said that what amuses me more is Stefan Wess talking about nightmares of AI - https://youtu.be/MaItaCQcYIE - Imagine machines learning from pessimistic human beings, humans have control, morals, sentiments towards certain things and can over power other humans incase of misdead. Some of the well established people like Elon Musk has the nightmares about AI’s negative advancements, and one of the reasons why the godfather of Tesla with other prominent personalties in tech industry have invested in OPEN AI a non-profit artificial intelligence research company - https://openai.com/blog/special-projects/ - As much as its fascinates me it worries me about our future too. But i beleive people who are not part of this tech maybe left behind and eventually have to jump in the AI tech bandwagon sooner or later. AI holds disruptive future of tech industry.

Computers --> Internet --> .com --> Mobiles --> Apps --> Chatbots --> Artificial Intelligence Bots

PS: Sorry, I couldn’t hyperlink it as I was replying from mobile and please bear with my englsih too.


Just wished my cousin sister. Bangalore. IISc report
#4

Tha’s great. Please invite them to join startups forum. We need all the inputs we can get from brilliant and enterprising folks to drive this forum towards a great community experience.[quote=“darshannchauhan, post:3, topic:6223”]
am very much mesmerised by Jibo
[/quote]

First time im hearing about it, and it looks totally awesome, disruptive and a tad bit dystopian too, when you think about what it means to the word family :smiley:
Thanks for sharing Darshan. Any clue what are the technologies involved in this ? I mean not at a 10000 feet level, but at a ground level ?

On a side note, since we are talking about intelligent machines in general , have you seen the last years viral video of the four legged MIT Cheetah Robot that can do a running jump ? The interview with the research team and the MIT NewsLetter also make for an interesting read.

It just goes to show that Indian Universities need to keep pace with what is going on at MIT and Cornell, else we too would be left behind in the pursuit of technological prowess, which powers our future. Do you know of any such interesting work in the field of Machine Learning going on at IITs or any of the technology related institutions / universities in our country ? Please do share if you do, so folks on this forum too can be inspired to think out of the box.

Yes and thats exactly what should drive a conversation around this on startups forum. We must be well aware of what is going on in this field, so that it does not come as a surprise to us twenty years later, when we realise robots are taking over the planet, and making less technically adept human beings redundant!
Do you have any suggestions how to approach this problem of dystopian proportions ?


#5

Guess What Ron? Even Jibo was made By MIT’s Robotics master Cynthia Breazeal.

By technology do you mean on what platform it was made on or coding was used to create? I absolutely have no idea as I am not a coder or engineer of any sorts

But if you browse around here maybe you would know! If you get to know do let me know I would curious to know too.

I read it somewhere, there are only 2% of AI Startups in India as compared to the US. Most of the startups in India are busy with their next round funding.

As of now nothing that I know of but do you remember the very famous and notorious IIT dropout and founder of Housing.com, Rahul Yadav. After resigning from Housing.com he started AI, AR & VR startup for government Intelligent Interfaces that used these technologies to expedite the work and he had quoted that if government made rental agreements online than government would save Rs 2,000 - 3,000 cr but unfortunately working with government is not easy thing and eventually he had to scrap the startup. He was backed by very prominent investors Mukesh Ambani, Micromax Founder Rahul Sharma, Paytm Founder Vijay Shekhar Sharma, Flipkart founders Sachin and Binny Bansal and Yuvraj Singh.

For now, it’s really tough to say since we have yet to explore AI with its self-learning capabilities, maybe AI bots can be uploaded with morals, sentiments, good behaviours but the real worry is that how tough the cyber-security system or firewall you create there is a hacker guy who will hack into it. I fear that maybe someone will hack into my robot to kill me. That’s my nightmare. So I am really looking forward to AI cyber security that Open AI are working on.


#6

thanks for that snippet i think we have found a common role model then !
Interesting work is always interesting :smiley:

The thing i know is that there are always two aspects to studying any subject,one is the science of it, and the other is the tools related to the science. I cant really say if Jibo SDK would give us an indepth idea of how the underlying science of AI (assuming it is classical AI) works, but yes it might give a general idea.The SDK tools might give an idea of programmable options related to AI and response to external stimulus, but i think one has to dig much deeper into the scientific principles , which is converted to notational maths and then proceed to understand the hardware innovations that are driving AI, to understand what exactly gives it a lovable personality.
Definitely will share anything on the AI front that i can dig up.

Amazing how the guy convinces investors. Wish i had half his panache! :smiley: But seriously why is he trying to use AI to work with governments . The digital narrative of our country is so nascent our govt websites, dont even have proper UIs , forget AI :smiley: If he really wants to solve government problems, he should try creating a better IRCTC, esp. to handle load balancing 10,000 web requests a second during tatkal timing :smiley:

[quote=“darshannchauhan, post:5, topic:6223”]AI bots can be uploaded with morals, sentiments, good behaviours but the real worry is that how tough the cyber-security system or firewall you create there is a hacker guy who will hack into it. I fear that maybe someone will hack into my robot to kill me. That’s my nightmare.
[/quote]
Inspired by you, ive spent the whole day digging up Machine Learning concepts, and im trying to come up with a theoretical model for machine learning. Long way to go but very much on the lines of what you are saying is a defense against the dark arts, or rather a behavioral simulation program that can make right / wrong decisions as opposed to left - right decisions.( i hope that cliche is easy to understand, i intend to use it now as a sales pitch :smiley: ).
I mention theoretical coz i dont think at this stage anything practical can be conjured so easily. Research into machine learning can easily take up to a year for any normal individual, before you know exactly what we are dealing with.

I think its excruciatingly difficult to come up with a cybersecurity and AI product idea, since both fields are supremely challenging and especially Cyber security the domain is the toughest i can think of personally to enter. Its easy to do an ethical hacking course by paying a couple of lakhs, and call yourself a white hacker, but seriously if you want to get into the field you need a good 15-20 years of experience in networking, programming and technology in general.

I will definitely share my project pitch with you for your feedback on Machine Learning if i can get it ready in time for the upcoming startup conferences in Bangalore.
For some reason, these conferences always happen in the latter part of year.


#7

Hi @Ron,

Busy with work and have not been able to reply but I will soon.

I found something on JIBO that might be interesting to you. http://stackshare.io/jibo

Do check out and let me know!


#8

Maybe with the power of his backers he taught he could have made difference but clearly it did not help!

Sounds interesting @Ron if your idea is persuasive enough you can go to Y combinator who will be willing to fund these kind of ideas as the worlds most powerful and leading incubators. [quote=“Ron, post:6, topic:6223”]
Its easy to do an ethical hacking course by paying a couple of lakhs, and call yourself a white hacker, but seriously if you want to get into the field you need a good 15-20 years of experience in networking, programming and technology in general.
[/quote]

I think you need to know basic idea to tackle and a vision to guide your team of coders to acheive your goals. Do you know Steve jobs never coded a single line, and today its top most valueable company in the world with very powerful OS. So if you have the vision you need a right mentor and right people to work for you!


#10

What do you do in AI can you elaborate here @suresh ?


#11

AUGMENTED AI BASED DATA ANALYTICS
GRMC supports you to increase the productivity and extend the life-period of your transportation assets, associated with road, rail, waterway or bridge networks. Our predictive analysis system provides maintenance information which is helpful to optimize performance in an effective and profitable way. We support you to make better decisions about maintenance of other assets in Telecom, Electric Utility, Defense, National security, Oil and gas etc. The targeted task can be easily executed with focus onto that particular asset. We increase the safety, performance and reliability in various verticals. Our solutions help you to plan, design, construct and operate utility infrastructure to meet ever-changing industry requirements, reduce costs and deliver sustainable, safe, timely, and reliable service. Our offerings for planning and design include mapping and geospatial analysis, site analysis and design, substation design, network design and management, transmission tower design and so on.


#12

@Ron “Elon Musk announced on Twitter that he is “making progress” on his neural lace design, which is designed to augment human intelligence and thereby avoid the inevitable takeover when A.I.s become smarter than us.”

Musk :raised_hands:


#13

Personally i dont think we should be celebrating Musk this early.
From the link you shared, the cortical interface (which he terms neural lace) he is designing to overcome the current limitation of using keyboards and hands to interact with Robotic AI is uncharted waters, and hence risky enough for us to weigh pros and cons before showing real technical interest in its API.
We already have machines like Deep Blue that are smarter than the smartest humans around, so as far as fascination with ultra intelligence is concerned, should it not be relative to the amount of good it is doing for humanity ?

Coming back to the video, so instead of Jibo becoming the pet, we will become like a house cat, due to the rate of advancement in AI .
Because of the implications it has for influencing decision making , I think ultra intelligent AI will soon be as dystopian a concept as the potential of nuclear weapons falling into the wrong hands.
So we must be careful, and understand this more, rather than just looking at it as something you want to learn and then pitch at a startup conference thing.

Thanks for sharing :wink:


#15

Very interested in this. Would love to know more about Indian AI Startups.


#16

Technologies you can use like python, R, machine learning


#17

We are into some serious AI. Have a look at Curl Analytics.