The limitations of neural networks

Today neural networks dominate the landscape of AI and AIOps, but Ive said many times that this is unsustainable. Neural networks have peaked in their ability to deliver effective and meaningful results. The science has issues with basic intractability, mismatch and inherent latency. Even though there is a lot of investment in neural networks, its bearing on AIOps and the real-time business community is limited. Which brings me on to computational neuroscience, which I believe will benefit AI enormously.

As I gaze into the future in terms of how AI is likely to evolve, I expect there to be a lot of crossover with computational neuroscience. Whats happening at the moment with neural networks is an early attempt to cross fertilize with AI, but this is failing and will continue to do so.

Its an attempt to take the complex and poorly understood behaviours of the human brain and associated nervous system and develop both mathematical and algorithmic models to try to understand their behaviour. You can compare computational neuroscience to economics or climate science. In all of these cases you have an immensely complex system with visible, but poorly understood, contours. We hope we can learn something about these systems to make high level predictions, which is achieved by building computational models that are either straight algorithms or a set of mathematical equations to try and get some insight into these large complex systems. This approach is entirely different from other scientific endeavours such as physics and chemistry, where you start with well defined behaviour and then try to build from the bottom up to understand, for example, why atoms behave the way they do, or how molecules or cells interact. Think of computational neuroscience, economics, and climate science as top down sciences, as opposed to classical bottom up sciences. Generally, computational neuroscience will give you some indication as to how the brain and nervous system works.

When you look at it that way, one of the things that becomes very interesting is that AI and computational neuroscience have many similarities. However, there is a perception that there is a massive difference between the two disciplines, many perceive computational neuroscience as a bottom up science and see AI as an engineering project. That is wrong, both of them are top down sciences that are investigating very similar and heavily overlapping domains. Therefore, in the next five to ten years we are going to see more crossovers between the two disciplines.

Firstly, there will be an increasing focus on how AI algorithms interact with one another. I think in most academic research and industrial efforts there is a lot of emphasis on developing and working with individual algorithms, but there is very little attention given to how the collection of algorithms interact with one another from an architectural perspective. One of the reasons why is because we naturally think of intelligence as a space that co-exists and there is no interacting structure. The truth is that algorithms need to be carefully choreographed with one another. This is very evident in the field of AIOps and how the Moogsoft platform has evolved. We have different types of algorithms which function at different times and hand off their results to one another. The result is very similar to the architecture off the human brain as we understand it.

As AI is deployed more systematically across more systems, the need to choreograph the interactions between the different algorithms will become more pressing. There is a vast body of knowledge which already exists on how, for example, visual systems interact with higher level conceptual categorization systems or how visual and auditory systems interact with one another. Therefore, it would be natural to look at the architecture of the brain as a starting point to design an optimal architecture for the interaction of various AI algorithms.

SEE ALSO:Data recovery: What matters when disaster hits

Secondly, AI research and industrial deployments has always focussed on centralized AI algorithms. In general, there is a drive to pull data in from various parts of the environment and take it to a single place where the AI algorithm is applied to it. I think increasingly there will be a focus on distributing algorithms geographically.

If you look at the way cognitive processes are enacted in the brain, and especially in the nervous system, it is evident it can become a model for how intelligence can be modularized and distributed not only conceptually but physically across a system. I think the way in which models are being developed on the computational neuroscience side to reflect distribution of intelligence will end up being a body of teachings for AI. To be fair, even in the field of computational neuroscience there has been insufficient focus on the need to modularize and distribute algorithmsbut its definitely coming.

Thirdly, as industry becomes more and more interested in robotics (the application of AI to automation) there will be an increased focus on how intelligence and AI algorithms interact with physical world processes. So, as robotics moves from being theoretical to a genuine industrial endeavour, the models that have been built to understand how the brain interacts with the nervous system and the external world will play an increasing role in the advancement of AI.

SEE ALSO:How to implement chatbots in an industrial context

Lastly, when we talk about machine learning or neural networks the focus is very much on the learning that takes place within an individual algorithm. It is not focused on how an entire system of algorithms evolves. As AI begins to recognize the importance of architecture and the choreography of algorithms; as it becomes more focused on distributed intelligence; as it becomes more focused on interacting with the external world; then I think were going to develop systems whose entire cognitive apparatus evolves and learns with time.

Computational neuroscience has absorbed and modified work conducted around cognitive psychology which has been embraced by the neural science world. I think this research has a lot to teach AI around the cognitive architectures it seeks to deploy in the industrial world.

These are the four big developments which will occur over the next five to ten years. Lessens learnt and models built in the computational neuroscience world will enter the world of research and industrial AI. As AI becomes more involved in business process execution, it starts to behave more like the brain and nervous system and hence its not a surprise that the work that has been done in computational neuroscience is likely to impact AI in the years ahead.

Read more:
Why computational neuroscience and AI will converge - JAXenter

FENS currently represents 43 European national and single discipline neuroscience societies with more than 20,000 member scientists from 33 countries across Europe. FENS promotes neuroscience research to policy-makers, funding bodies and the general public, both regionally and internationally. Hence, FENS promotes excellence in neuroscience research and facilitates exchanges and networking between neuroscientists within the European Research Area and beyond. We spoke to FENS President, Professor Carmen Sandi, about some of the latest trends and challenges the neuroscience sector is currently facing.

Neuroscience is a very broad and vibrant discipline aimed at understanding how the brain works and which mechanisms underlie its different types of dysfunction, where new topics and techniques are constantly added. In recent years, the emergence of possibilities for large data collection and management, and advances in computational sciences, have transformed the way neuroscience is done today and how it relates to society.

I should mention the renewed emergence of Artificial Intelligence (AI), which at its base is a neuroscientific discipline, and how it relates to our day-to-day activities. Similarly, technological advances in imaging and molecular and genetic tools have transformed the way we, as scientists, can understand complex behaviours and dissect the pathways and causes underlying behaviour and cognition as well as the debilitating diseases of the nervous system. I am very hopeful that a decade from now, insights derived from todays investments in neuroscience research will form a new basis for providing early diagnosis, prevention and in some cases cures, for serious diseases that currently represent an unparalleled burden in Europe.

As the voice of neuroscience in Europe, FENS promotes collaboration, networking and career development for the neuroscience community at large, with a particular focus on European early-career scientists. FENS supports education and training by providing information, organising schools and training programmes, and offering grants. FENS facilitates the dissemination of scientific information through its journal, the European Journal of Neuroscience. It also encourages interactions between neuroscientists and related scientists within and outside Europe by holding scientific meetings such as the biannual neuroscience conference series The Brain Conferences and the biennial FENS Forum.

The FENS Forum of Neuroscience is FENS flagship event. It brings together more than 7,000 international neuroscientists every second year and fosters scientific exchange and interdisciplinary collaboration. It attracts scientists from academic, fundamental sciences to preclinical and pharma scientists who are working on translational understanding of the mechanisms underlying brain diseases. Facilitating the exchange between these diverse and international scientists is an important objective of the Forum, and for FENS more generally. The next Forum will take place in Glasgow on 11-15 July 2020.

The Neuroscience Schools are part of FENS broader higher education and training strategy, which offers a wide range of opportunities in continued education and career development. At the end of your university education or when you start as a postdoc, when dedicated training in your particular field of neuroscience is often limited, formalised training such as that offered by FENS is extremely important for the success of a young scientist today.

The FENS schools are run within an intimate setting by leading scientists around novel concepts and theories in neuroscience. This format facilitates both formal and informal exchanges and leads to a high degree of interaction between the students and the faculty. FENS offers a wide range of training opportunities. The Cajal Advanced Neuroscience Training Programme for instance, which is a new training concept in Europe, offers a series of three-week hands-on training courses on timely neuroscience topics. The Cajal programme, which has been developed by FENS and IBRO and is supported by the Gatsby Foundation, is a unique platform for dedicated hands-on training across most disciplines in modern neuroscience. Tied in with these training programmes, FENS has also developed a Network of European Neuroscience Schools that brings together most of the Universities and Centres that are offering graduate training in neuroscience. The network, which currently represents more than 180 programmes, was recently expanded to also include online courses and programmes. Through this network, FENS can act at the European level to coordinate and influence neuroscience education.

Our mission at FENS is to advance research and education in neuroscience and, within this context, to promote neuroscience research to policy-makers, funding bodies and the general public. Too often, the societal value of investing in fundamental research is underestimated. The challenge to FENS thus lies in demonstrating the importance of investing in knowledge-generating fundamental research and showing how it fuels applied research and innovation. Promoting interaction and coordination between neuroscientists throughout the value chain is probably our largest challenge within the scientific arena of today, as well as connecting European research with efforts outside of Europe. To that effect, FENS regularly organises meetings and events, where scientists can connect across research areas and from different ends of the value-generating research enterprise. The FENS Forum, for instance, is a great platform to showcase innovation: delegates can learn more about new technological developments and meet with their peers, from basic to translational science. We also conduct a series of outreach activities to promote the understanding of neuroscience among the general public and decision-makers as well as coordinating events and activities within the scope of the European Brain Council (EBC) to influence how brain related policies and priorities are defined.

In Europe alone, an estimated 179 million people live with brain disorders and the estimated cost of these disorders in Europe since January 2019 has exceeded 615bn. Brain health, and by extension, brain research, need to be identified as funding priorities by the European Union and across its Member States. Policymakers, at both national and European scales, need to ensure that research remains a priority in order to secure new ideas and development of new technologies.

FENS has developed an advocacy strategy that comprises three levels. At the national level, FENS engages with policy-makers through its national neuroscience member societies: they identify and act on the specific needs for neuroscience advocacy. Under the umbrella of the EBC, FENS, together with other member organisations, regularly interacts as an advisor to European institutions and provides expertise and recommendations. Speaking as one voice towards the European Institutions, the EBC stands as a unique platform to foster cooperation between its member organisations and other stakeholders. At the global level, FENS works with other leading organisations to raise public awareness and promote investment in and cohesion of neuroscience research. In this context, FENS provides support for advocacy and outreach programmes across the globe for the dissemination and support of brain research.

2020 is a Forum year! The 12th edition of the FENS Forum will take place on 11-15 July in Glasgow, UK. It has a high-quality and wide-ranging scientific programme, designed to showcase the frontlines of science, giving the floor to renowned speakers and the most up-to-date discoveries and innovation. The Forum is always a great platform to discuss science, regardless of career stage. On a slightly longer horizon, I believe that FENS will continue to play a vital role in coordinating knowledge exchange in Europe and globally. I am confident that the scientific community will continue to embrace the values that are provided by scientific societies in Europe as I truly believe FENS and similar organisations represent the best channels to support the delivery of tomorrows knowledge and cures.

Professor Carmen Sandi

President

Federation of European Neuroscience Societies

Tweet @FENSorg

Home

Recommended Related Articles

Link:
Promoting neuroscience in Europe with FENS - SciTech Europa

Talking Heads frontman David Byrne has been interested in the intersection of art and science for years. When hes not debuting a Broadway show, launching an online magazine or creating a show about Joan of Arc or Imelda Marcos, hes exploring how neuroscience plays with our perceptions.

Theater of the Mind, to have its world premiere in Denver in August 2020, picks up on themes Byrne explored in a Silicon Valley art exhibit in 2016. That exhibit, The Institute Presents: NEUROSOCIETY, featured a series of interactive environments created in conjunction with PACE Arts + Technology in Menlo Park, California, that questioned human perception and bias.

The goal of Byrne and writer Mala Gaonkar for Theater of the Mind is to blend sensory experiments with theatrical entertainment in a seamless show.

Think of it as a carny hall of mirrors 2.0 with input from neuroscientists. An intimate audience of 16 people at a time will move through specially designed environments throughout 15,000 square feet within a warehouse. Sometimes wearing VR headsets. With a storyteller guide.

The immersive show will debut here, thanks to the Off-Center branch of the Denver Center for the Performing Arts, which has gained national attention for its exploits in immersive theater.

Charlie Miller, Off-Center curator, heard from a friend that the NEUROSOCIETY show was mind-blowing and last year approached Byrne about a collaboration. This is a world premiere of a project David has been developing for four to five years. There have been earlier workshop iterations and it continues to evolve with major script revisions.

Part of what is exciting about it for me as a producer, Miller said, is were still figuring out how the machine of it will run. Part of the reason were not sharing information about scheduling is that were still figuring out how frequently groups can move through. Its a very high-tech undertaking in terms of the technology in the experiments and in the technology thats used to run the show.

Due to the small size of each audience group, the nightly or weekly total wont be huge, Miller said, but enough to bring in revenue to make it economically feasible. Ticket sales only cover a portion of the overall cost. As a nonprofit, Off-Center relies on grants, SCFD funding and investment from DCPA as a whole. The balance is better than the one-on-ones we did last year (like Between Us), which were very heavily grant subsidized.

In the early version, NEUROSOCIETY viewers were led through several rooms. In one they saw their hand grow to giant size and observed themselves embodied in a dolls body. In another they saw moving objects freeze, and witnessed complete darkness and some single very bright flashes of light (not strobes). The installations and the script have changed substantially since then, Miller said.

The long-term hope is that Off-Centers name will be attached to this project when it is mounted in other cities, gaining further prominence. DCPA will be involved if it can go on to a life in other cities which we all hope it can. Our main focus is to get this thing off the ground in Denver. If we get it right, that will open doors for the future.

But Theater of the Mind will not be easy to transport elsewhere, given its specific requirements. There are close to a dozen different environments were building now that the audience moves through, a whole VR experience being custom built by a series of technology companies, it requires custom software. It will be a challenge to find space of a certain size that can accommodate it, Miller said.

He wont hint at the Denver location of the historic warehouse where the show will be produced. In size and scale, the installation will be similar to Sweet & Lucky, the inventive and engaging immersive show commissioned by Off-Center and created with New York-based Third Rail Projects in 2016. (The warehouse leased for that production, behind Mission Ballroom, is no longer available.)

The five to 10 experiments included in Theater of the Mind are meant to demonstrate how easily our minds can be tricked. The show engages all of the senses, including taste (expect a number of disclaimers and FAQs beforehand).

The team is passionate about being honest with the audience the magic is in the science, its not in theater magic, Miller said. Were not playing tricks on people. What happens in your brain is really the magic there. Its very true to science.

In fact, there is a position on the creative team director of technology helping the theater folks preserve the integrity of the science.

We always go back to whats going to make the science most effective. Thats the starting point. The purity of the science, Miller said.

Expect ancillary programs to include the local scientific community once the show is up and running.

Specifics on dates, times and tickets are not yet available but signing up on the shows website ensures those who are interested will hear more as information is released.

This reporting is made possible by our members. You can directly support independent watchdog journalism in Colorado for as little as $5 a month. Start here: coloradosun.com/join

Go here to read the rest:
David Byrne is building a neuroscience-powered hall of mirrors somewhere in Denver - The Colorado Sun

It's the latest fad in Silicon Valley. By reducing the brain's feel-good chemical known as dopamine cutting back on things like food, sex, alcohol, social media and technology followers believe that they can "reset" the brain to be more effective and appreciate simple things more easily.

Some even go so far as avoiding all social activities, and even eye contact.

The exercise, dubbed "dopamine fasting" by San Francisco psychologist Dr Cameron Sepah, is now getting increasing international attention. But what exactly is it? And does it work?

As someone who studies the brain's reward system, I'd like to share my knowledge with you.

Dopamine is a neurotransmitter a chemical messenger produced in the brain. It is sent around the brain conveying signals related to functions such as motor control, memory, arousal and reward processing.

For example, too little dopamine can result in disorders like Parkinson's Disease, involving symptoms of muscle rigidity, tremors and changes in speech and gait. One of the treatments for Parkinson's is the drug L-DOPA, which can cross the blood-brain barrier and be converted into dopamine to help ease the symptoms.

Dopamine is also important in the reward system in the brain. It is activated by primary rewards like food, sex and drugs.

Importantly, the brain's reward system can "learn" over time cues in our environment that we associate with potential rewards can increase the activity of dopamine even in the absence of an actual reward. So just being in a sweet shop and thinking about sweets can activate our brain's dopamine.

Dopamine transporter activity in the brains of a control and methamphetamine abuser. (National Institute on Drug Abuse)

This expectation and anticipation of rewards is called the "wanting" in neuroscience language. As one of the main symptoms of depression is "anhedonia" the lack of wanting, interest and pleasure in normally rewarding experiences dysfunctional dopamine regulation has also been linked to this disorder.

Some treatments for depression, such as the drug bupropion, are designed to increase dopamine levels in the brain.

So, given the important role of dopamine in vital functions in the human brain, why would we want to fast from it? The idea of dopamine fasting is based on the knowledge that dopamine is involved in unhealthy addictive behaviours.

As described, dopamine underpins wanting. For instance, a drug addict may say they no longer want to take drugs. But when in certain places where drug-related cues are present, the brain's wanting system kicks in and addicts are overcome with strong urges to take the drug.

Dopamine fasters believe that they can reduce desires and craving for unhealthy and even unwanted behaviours by reducing dopamine.

First we need to be clear, it is certainly not advisable, even if we could, to reduce the amount of dopamine in the brain as we need it for everyday normal functions.

Further, simply banning a particular reward, like social media, isn't going to reduce the levels of dopamine per se, but rather it can help reduce the stimulation of dopamine.

Therefore it is possible to reduce the amount of dopamine activity. But the key to doing this is to reduce our exposure to the triggers associated with the rewards that initiate the wanting for the rewards in the first place.

After all, it is these cues that initiate the craving and the desires to engage in behaviours that help us get the rewards. Thus just cutting out rewards doesn't necessarily stop the brain from making us crave them activating dopamine.

However, that this would "reset the brain" is not really correct there is no way of even knowing what the baseline is. So from a neuroscience perspective, this is nonsense for the time being.

If you find that you want to cut down on what you feel are unhealthy behaviours, such as spending too much time on social media or overeating, then you could start by reducing your exposure to the environmental cues that trigger the desires to carry out the unhealthy behaviours.

For example, if you go on your phone too much in the evenings when you are alone, try turning off the notifications sounds. This way dopamine is not being activated by the cues and therefore not signalling the urges to pick up the phone.

And if you think you drink too much alcohol ending up in bars with work colleagues most nights of the week try to go somewhere else in the evenings, such as the cinema.

The symptoms of unhealthy behaviours are similar to the signs of substance abuse. These might include spending the majority of the time engaging in the behaviour, continuing the behaviour despite physical and/or mental harm, having trouble cutting back despite wanting to stop and neglecting work, school or family.

You may even experience symptoms of withdrawal (for example, depression, irritability) when trying to stop.

In these instances, you may want to think about removing the cues that stimulate your dopamine neurons a sort of dopamine fasting.

Ciara McCabe, Associate Professor, University of Reading.

This article is republished from The Conversation under a Creative Commons license. Read the original article.

See the article here:
'Dopamine Fasting' Is Silicon Valley's Latest Trend. Here's What an Expert Has to Say - ScienceAlert