#DeepSeek Decoded: What Educators Must Know Today
Screenshot taken from DeepSeek by B.Clarke

#DeepSeek Decoded: What Educators Must Know Today

It had to happen sooner or later…a contender in the Gen AI space. This week Chinese developed AI tool #DeepSeek has topped the app charts. I for one am super excited. This means that the Open AI ChatGPT monopoly will be challenged and surely that is a good thing? Monopolisation in my opinion does not lead to growth and it is growth that moves us all forward, so bring on contenders.

With my educator hat on, where may you start in the classroom to discuss this topic with your learners? Here are 4 things you need to know about #DeepSeek

#Deepseek works on reinforcement learning, which in its simplest form is a trial and error correction model. See my work here ( lesson 5 – 7) for educators to help with understanding the concept of reinforcement learning and “Teaching about AI

  1. We haven’t come to “trust” #DeepSeek as yet. Trust has to be built. There is lots of talk around safety concerns and that possibly we may be “spied upon” through usage. In truth can we truly trust any of the “free” products that we all download on a daily basis? Remember nothing is free without a good reason. For me the jury is out on that one!
  2. It is apparently more energy efficient, which if this is true will be better for the environment. Think GreenIT. As I have said many times tech is part of the problem and part of the solution to the climate crisis. See my work here on embedding GreenIT teaching into Computing/tech lessons
  3. The training costs are less which means it is a possible game changer and we can see smaller companies and developers get involved as they don’t have to be overly concerned about big budgets. This may be the start of more equitable tech. An issue that the sector has been battling for decades. What do you think?

?As we continue through the fourth Industrial Revolution, we are a long way from the first industrial revolution and steam power. Let’s see how the continued AI disruption unfolds. Are you going to be using #DeepSeek ? So far from my testing, it seems accurate up to the point it was trained which was just over six months ago.

What are your concerns OR are you all in with the possible opportunities?


Here are a few ways in which you can work with me.

I think the biggest concern I have with DeepSeek is that it’s already proven to cleanse or omit data that may make China look bad. Ask it what happened at Tiananmen Square and it won’t give you the answer. This points to the flaws in how data can easily be manipulated, which highlights that now more than ever questioning the data/answers you get from these tools and ensuring their accuracy is imperative.

要查看或添加评论,请登录

Beverly Clarke MBE的更多文章

社区洞察

其他会员也浏览了