Mastodon
Right, so artificial networks also need sleep!

Right, so artificial networks also need sleep!

by | 2022 Nov 21 | Public, Quick Hits

Quick Hits
Daily brief research updates from the cognitive sciences


Ishould say that it is artificial neural networks that seem to need sleep, or a rest. But isn’t one point of an artificial network that that of it not needing sleep, or rest?

For those of you who don’t know, artificial neural networks are networks built by engineers in the computing space to mimic the brain’s cells and therefore hope to get better computing, or different, computing outcomes. Therefore, it comes as a real surprise that, apparently, rest improves their performance. It comes as a real surprise because these are not biological entities, there are numerous reason we need rest as human beings, as biological beings. One is that there is a slow build-up of toxic material as we function. Another is that there is also constant genetic damage that needs to be repaired.

So why do these networks need rest? First let’s understand what happens. These networks are designed to mimic neuronal functions – so far so good – and they have become really good at some things such as computational speed. But there is something called catastrophic forgetting – no, not like when we stand in the supermarket and can’t for the life remember what we wanted. This is when these networks learn sequentially, one task after another, new information can then overwrite old information and it is gone, “forgotten”.

Golden et al. at the University of California have now reported that when these artificial networks are trained on new tasks but with periods off-line mimicking sleep, they could replay old memories but without using old training data. This is due to the patterns that are replicated during our biological sleep. In our sleep the synapses, connections that is, between neurons are strengthened. You brain essentially replays your day’s input and strengthens memories during sleep (that’s one of many reasons sleep is so important). When this process was replicated it mitigated this catastrophic forgetting.

So, fascinating it is that we are building artificial neural networks that replicate the brain’s processes for better computational processing power but also fascinating that these artificial networks also improve performance with sleep.

This goes to show that good old biology, and evolution, seems to have got it right. And for us also another reminder of the importance of getting a good night’s sleep.

Quick Hits Alerts
Article Alerts
Newsletter
Premium Brain
Andy Habermacher

Andy Habermacher

Andy is author of leading brains Review, Neuroleadership, and multiple other books. He has been intensively involved in writing and research into neuroleadership and is considered one of Europe’s leading experts. He is also a well-known public speaker, speaking on the brain and human behaviour.

Andy is also a masters athlete (middle distance running) and competes regularly at international competitions (and holds a few national records in his age category).

twitter / LinkedIn

References

Ryan Golden, Jean Erik Delanois, Pavel Sanda, Maxim Bazhenov. 
Sleep prevents catastrophic forgetting in spiking neural networks by forming a joint synaptic weight representation.
 PLOS Computational Biology, 2022; 18 (11): e1010628
DOI: 10.1371/journal.pcbi.1010628

More Quick Hits

Making Voting More Effective for Better Decisions

Making Voting More Effective for Better Decisions

by | 2022 Nov 17 | Public, Quick Hits

Quick Hits
Daily brief research updates from the cognitive sciences

Voting is something many of us do regularly as adults in different ways. But the most common way is plurality voting – that is one voter for one person that can only be cast once.

However, there are different ways to vote. Popular shows such as American Idol use multivoting whereby the audience memebrs have 10 votes to share as they see fit. They can use all ten for one candidate or distribute them as they wish across candidates.

Another method is ranked-choice whereby choices should be ranked in the order one sees as best.

The question though is what is best? This is the question researchers at the University of Washington wanted to find out but specifically into what enables better decision making. In many decision-making bodies such as in business or government simple plurality voting is used.

To do this the researchers used 93 teams of graduates in an anti-terrorism exercise modelled on teams used post 9/11. 31 teams each used the different voting methods. They first read information on three suspects and voted on who they thought was the greatest threat – they then discussed the suspects after the vote, and then voted again.

In the plurality voting scenario, teams identified the correct suspect only 31% of the time. Basically, around about pure chance. So actually ineffective. Did the other voting scenarios perform better?

Well, ranked-choice performed only a little better with 32% of the time. But the largest difference was in multivoting – here 45% of teams identified the correct suspect. But there was another surprise.

This was that multivoting seemed to be more effective before the discussion. We would assume that voting and then discussing would lead to better insight and therefore making a better decision. But this was not the case. In multivoting it seems that it forces people to think through things more carefully in the first place. The ensuing discussion can then side-track thinking for multiple reasons.

So, according to this research at least, multivoting is easily the most effective method for making decisions. Something many businesses should seriously consider – particularly in important decisions – and prior to discussion!

It could be too complicated for political voting but then again, I live in Switzerland, and we already use a similar version. Implementing this in a different country would be a challenge – but for now, for decision-making, we do know that multivoting appears to be the best method. Well, for what it’s worth, it has my vote(s).

Quick Hits Alerts
Article Alerts
Newsletter
Premium Brain
Andy Habermacher

Andy Habermacher

Andy is author of leading brains Review, Neuroleadership, and multiple other books. He has been intensively involved in writing and research into neuroleadership and is considered one of Europe’s leading experts. He is also a well-known public speaker, speaking on the brain and human behaviour.

Andy is also a masters athlete (middle distance running) and competes regularly at international competitions (and holds a few national records in his age category).

twitter / LinkedIn

References

Michael D Johnson, Eli Awtrey, Wei Jee Ong.
Verdicts, Elections, and Counterterrorism: When Groups Take Unofficial Votes.
Academy of Management Discoveries, 2022
DOI: 10.5465/amd.2021.0099

More Quick Hits

We’re Bad at Remembering How Happy We Were

We’re Bad at Remembering How Happy We Were

by | 2022 Nov 16 | Public, Quick Hits

Quick Hits
Daily brief research updates from the cognitive sciences

The good ‘ole days, eh! There is some evidence to suggest that we always view the past though rose-tinted glasses – feeling that the past was somehow better than the present. However, research just out has shown that that doesn’t seem to be the case.

Research by Alberto Prati, Claudia Senik for the Association of Psychological Science has shown that current feelings impact our view of the past. To do this they have assessed data from different longitudinal studies into how happy people feel over time. These include a German study over 10 years of 11’000+ participants, a British study over 12 years of 20’000+ participants, a French study of over 18’000 participants, and about 4’000 results from the USA over a period of 35 years.

What did they find?

What they found is that there seem to be some mix up between current ratings and change in happiness over time. This in contrast to the past is better hypotheses. This showed that those who felt happy now rate their past happiness as lower than it actually was. It seems that feeling happy now suggests an improvement or a contrast to the past and so the past must have been less good.

In contrast the opposite happened with those with current lower happiness – they rated their past happiness as higher than it actually was.

This shows that our past memories are influenced by our current states. Prati and Senik plan to further research how memories impact current life decisions – that will be interesting to see. I also wonder how this influences political climate!

Quick Hits Alerts
Article Alerts
Newsletter
Premium Brain
Andy Habermacher

Andy Habermacher

Andy is author of leading brains Review, Neuroleadership, and multiple other books. He has been intensively involved in writing and research into neuroleadership and is considered one of Europe’s leading experts. He is also a well-known public speaker, speaking on the brain and human behaviour.

Andy is also a masters athlete (middle distance running) and competes regularly at international competitions (and holds a few national records in his age category).

twitter / LinkedIn

References

Alberto Prati, Claudia Senik.
Feeling Good Is Feeling Better.
Psychological Science, 2022; 33 (11): 1828
DOI: 10.1177/09567976221096158

More Quick Hits

100 Years of Research Reveal the Most Effective Methods for Learning

100 Years of Research Reveal the Most Effective Methods for Learning

by | 2022 Nov 15 | Public, Quick Hits

Quick Hits
Daily brief research updates from the cognitive sciences

learning memory brain

Studying online

 

We’d all like to be able to learn easily. Read something and remember it, listen in on a call and never forget anything, but we all know, well the vast majority of us, that it isn’t always that easy. Over the years and decades many practises have also been developed to help learning, ranging from learning in your sleep to meditative methods. But do these really help?

Well, this is what Shana Carpenter et al. of Iowa State University wanted to find out. To do this they reviewed and analysed over 200 studies ranging over 100 years to find some clear answers. And the results?

The results show that basically two strategies are the most effective and therefore the most important.

These are not sexy new techniques – in fact quite boring. They are spacing and retrieval practice. That’s it!

Spacing is the concept of spacing learning out into more bite-sized chunks. For example, in one study medical students received training on surgery training over three weeks vs. one intensive day. Those in the spaced learning group performed better one week after training had finished but also, importantly, one year later.

I have reviewed spaced, or punctuated, learning previously. I have also reported on brain processes and fatigue during mini learning and break phases (here and here).

The second technique is also a low tech, old-fashioned, and effortful: the technique of learning retrieval. This simply means trying to remember what you have learned. This is the high effort version and probably the one we also try to avoid, particularly when by ourselves. This is more effective than the easy method which is just rereading your notes or the textbook again. The important part seems to be the active retrieval part, actually making an effort to get it out again.

So, this on one hand is a bit boring, no new sexy techniques. It is also really important – two simple techniques will improve learning for anyone and anyone can do it!

So, if learning something new, space it out it bite-sized chunks, and make an effort to remember what you covered and learned.

That’s it, that simple.

Now let’s see if I can remember what I have just written…

Quick Hits Alerts
Article Alerts
Newsletter
Premium Brain
Andy Habermacher

Andy Habermacher

Andy is author of leading brains Review, Neuroleadership, and multiple other books. He has been intensively involved in writing and research into neuroleadership and is considered one of Europe’s leading experts. He is also a well-known public speaker, speaking on the brain and human behaviour.

Andy is also a masters athlete (middle distance running) and competes regularly at international competitions (and holds a few national records in his age category).

twitter / LinkedIn

References

Shana K. Carpenter, Steven C. Pan, Andrew C. Butler. 
The science of effective learning with spacing and retrieval practice. 
Nature Reviews Psychology, 2022; 1 (9): 496
DOI: 10.1038/s44159-022-00089-1

More Quick Hits

Why Too Much Talent May Harm Performance

Why Too Much Talent May Harm Performance

by | 2022 Nov 14 | Public, Quick Hits

Quick Hits
Daily brief research updates from the cognitive sciences

team talent performance

That may sound like an odd thing to say, and this research is now actually a few years old, but is nevertheless still misunderstood. We all seem to generally feel that more talent leads to a better team. This applies to sports teams as well as to business teams. I have just come back from a conference in the life sciences sector and one panel was talking about talent. The discussion revolved around finding the best people.

This research suggests, that finding the best people may not be the most important strategy but finding the most suitable people – that is different – but this also isn’t always the case. It depends on how the team must operate together.

Adam Galinsky and Vikram Pandit of Columbia Business School have researched a number of team-based situations, from egg production in chicken coop. Yes, you read that correctly, chickens also need to produce effectively, and their performance, egg laying, drops in certain teams! To 10 seasons of professional basketball and baseball.

What did they find?

They found that “team coordination suffers when there is too much talent”. Not a good thing for team performance. Simply put, too much talent creates a conflictual “pecking order”, see that research into the chickens is suitable, top talent tries, understandably to be higher in the pecking order and this causes conflict and inefficiencies and conflict.

This is similar to what I wrote about in my popular article on underperforming high performers, when some people can help teams to perform better without being star performers themselves.

This means that just stacking teams with talent is not necessarily the best strategy – but this is what we automatically do. Intuitively it feels like the best thing to do – whether in sports or business. However, Galinsky does note that it depends on the team. Or rather it depends on the amount of collaboration needed.

Teams that require higher collaboration, in Galinsky’s research basketball teams, need a range of talent levels to perform to their best. In contrast teams that need less coordination and can work individually, in this case baseball teams, pitchers and hitters work independently, stacking the team with talent is a good strategy.

So, there you go – in business or sports you do need to think of how teams need to operate together and therefore what sort of talent you need. Our own internal data with our assessments show the same thing. And yes, you can measure team cohesion.

But for now, be careful of how much talent you wish for.

Quick Hits Alerts
Article Alerts
Newsletter
Premium Brain
Andy Habermacher

Andy Habermacher

Andy is author of leading brains Review, Neuroleadership, and multiple other books. He has been intensively involved in writing and research into neuroleadership and is considered one of Europe’s leading experts. He is also a well-known public speaker, speaking on the brain and human behaviour.

Andy is also a masters athlete (middle distance running) and competes regularly at international competitions (and holds a few national records in his age category).

twitter / LinkedIn

References

Swaab, R. I., Schaerer, M., Anicich, E. M., Ronay, R., & Galinsky, A. D. (2014).
The Too-Much-Talent Effect: Team Interdependence Determines When More Talent Is Too Much or Not Enough.
Psychological Science, 25(8).
https://doi.org/10.1177/0956797614537280

More Quick Hits