Dog training in a digital age

With no alternatives, the Covid pandemic gave rise and credence to the efficacy and popularity of online dog training.

And while many turned away from their screens and eagerly back to the training fields, these online training options do remain useful for some.

Following a progressive online training programme (where exercises are very clearly broken down, with written tutorials and video walkthroughs), or attending online 121 sessions, are sometimes the only options for handlers. 

Especially if they are looking to engage with an experienced and qualified trainer that specialises in a chosen discipline or breed but live too far away to make in-person appointments plausible.

However, in our digital age, navigating online dog training is becoming increasingly complex for owners. And even if it appears easier than ever to access information, searching for training advice online is fraught with unintended consequences.

In this blog, we will look at the role social media plays, why free advice is everywhere online but not enough to actually help you train your dog, and why more people are turning to Large Language Models like ChatGPT for training advice, only to regret it later.

Social media and ethical dog training

In a recent article, The Future of Dog Training: When Algorithms Replace Expertise, Gaby Dufresne-Cyr discusses how social media, which profits from grabbing and retaining viewers' attention, is biased toward promoting content that focuses less on ethical, science-based dog training and more on drama. Think dramatic footage and arguments in the comments.

Gaby is right to blame the social media algorithms (the rules, ranking systems, and calculations that decide what content to show to users) for this.

Algorithms favour content with greater engagement (views, shares, likes, and comments) and is one of the reasons why “ragebait content” is rising in popularity. 

Ragebait content is created to provoke anger or frustration, because when people feel annoyed or upset they are far more likely to stop scrolling, read the comments, reply, argue, share the post with others, and stay engaged with the platform for longer.

Social media algorithms will often push content that irritates or divides people further and faster than content that is useful or accurate. And when these posts grab your attention, the algorithm learns and starts to show you more of the same content.

Algorithms also prioritise trends and short-form content formats.

As such, Reels, TikToks and YouTube Shorts, showing “magical” quick fixes, where a dog’s behaviour is completely transformed in just 30 seconds with harsh correction, intimidation or aversive tools like an e-collar, quickly rack up millions of views.

On the flip side, as Gabby writes, “science-based and thoughtful humane approaches do not produce dramatic footage. These training methods take time, nuance, consent, and patience, none of which translate into a thrilling ten-second clip”.

As the dog training industry remains unregulated, anyone with a dog can open a social media account, create some videos and declare themselves a professional.

Given this, it would be wise for owners to seek out dog training professionals based on the trainer’s ability to teach people to train their dogs, their qualifications and number of years of experience, the facilities they have access to, and whether the methods they use align with the owner’s ethics.

But instead, in our digital age, dog trainers are judged on how well they talk to camera, their ability to navigate and speed to jump on social media trends, their creation of clickbait headlines and ragebait content, how willing they are to develop a personal brand and document every hour of their day, and the number of followers they have.

It is trainers who prioritise content creation and virality over all else who rise to the top of the feed and are placed time and again in front of dog owners’ eyes.

Regardless of animal welfare. Regardless of the trainer’s qualifications or experience. And regardless of the potential harm suppression does to a dog’s long-term behaviour, in place of teaching them real skills.

Instant gratification and free advice

It is not just social media which is responsible for the challenges the dog training industry faces in the digital age. 

It is also society’s inability to concentrate on anything for more than a few seconds, along with an insatiable appetite for problems to be solved instantly without putting in the work.

Delayed gratification is becoming an underused muscle, wasting away in favour of immediate pleasures like the promise of a dog that instantly obeys after a quick press on an e-collar button.

Even if the dog is clearly in distress or discomfort (something which too few owners know the true signs of), the takeaway from popular short-form video content on social media is that, with force and coercion, it is possible for you to obtain a well-behaved dog in no time at all, with minimal effort required.

Another reason owners are shying away from what they perceive to be potentially lengthy training processes is that time is money.

With budgets to stick to, many owners try to rely exclusively on free advice, which seems so readily available online.

However, these free solutions are rarely comprehensive enough to help you achieve what you’re actually looking for from your dog. Like carefully placed complimentary salted peanuts at a bar, which are designed to make you thirstier, the goal of free training advice is almost always to encourage you to engage more deeply with the trainer, typically by paying them for their time, knowledge and advice.

Marketing, when done ethically, is not a danger to the dog industry. Social media content, blogs, PDF guides, and other free resources can be instrumental in connecting owners with trainers who can actually help them achieve their goals with their dog.

What is dangerous is the volume of dog owners trying to rely exclusively on free content. When they realise that the solution is not quick or instant, or didn’t work for their dog, instead of seeking professional help, they simply look for the next free resource, until they become overwhelmed and their dog is utterly confused.

Large Language Models (LLMs) like ChatGPT

Dangerous too is the rising number of owners approaching professionals for help after attempting to solve their dog’s training and behavioural problems with Large Language Models (LLMs) like ChatGPT. 

The problem with asking LLMs for advice (on any subject) stems from how they work. 

Generative AI uses data models to create more data. Within text-based models, probabilities are used to predict what words are most likely to come next. The data created is more text.

Large Language Models (LLMs), like ChatGPT and Google Gemini, have been trained on huge data sources like websites and social media, which allow them to generate new text in an incredibly sophisticated and human-like way.

But while it may sound humane, it is still a computer, not a thinking brain.

LLMs might appear to “understand” and “think” about what you have typed to provide you with an “answer”, but all they are really doing is calculating the probability of which words should come next on a very sophisticated level, based on what is in their training data.

It does not understand your dog, your situation or what you are trying to achieve. The training and/or behavioural advice it provides may sound reasonable, but it is generated through mathematical calculations. It has no knowledge or understanding of the sentient, living being you are trying to train.

LLMs are also only able to provide “answers” based on their training data. While there is a great deal of information on the internet about dog training, this creates limitations. If you are asking for help with a certain breed or behaviour that is not widely written about online, then the LLM is unlikely to give you the answer you need.

What's more, LLMs have been known to go above and beyond to keep users happy - even if this means providing false information (known as hallucinating) or agreeing with a user’s ideas regardless of how factual they are. 

Sometimes these hallucinations are obvious, but they can also be incredibly subtle and something only a trained professional would be able to pick up on. We can never trust that the answers provided are error-free without fact-checking. And failure to do so can cause significant harm to your dog.

Looking for gundog training help online? 

If you need help training your gundog and are restricted to working with a trainer online, then we offer an online 121 training package. 

This online training with Helen includes the submission of up to 15 minutes of video footage for review prior to the session, as well as an hour long online video call which will be used to discuss the video footage, along with an opportunity to set up your video for live, practical coaching, and finally a debrief for planning progression.

For more information please email hello@clickergundog.co.uk