A union can help.
Declaration of function Recently, Netflix has caused a stir among Hollywood screenwriters, who are currently on He hits. The position is called “Artificial Intelligence Expert”. He is responsible for the movie search algorithms at Netflix, with a salary of 9 million kroner. Average salary for screenwriters who make movies that algorithms are looking for 480,000 NOK. This is only 5 percent of what a Netflix expert would earn.
This difference illustrates a major theme in today’s discussion of artificial intelligence.
Due to the rapid spread of generative AI, anyone can ask a chatbot, such as ChatGPT, to write a movie script in the style of a successful and experienced screenwriter and get Relatively good project They can work on it.
The criticism of the scriptwriters is that the language model is trained on existing scripts that are available online. For experienced writers, it seems like theft when AI experts train language models on their texts without permission, and then launch services that help everyone write like them.
As the level of skill required to write screenplays decreases, competition for writing jobs increases – and wages decrease.
This case provides an interesting insight into how AI will affect working life and the requirements of knowledge workers for negotiating wages in the coming years.
AI companies like OpenAI and Google have built services like ChatGPT by gathering vast amounts of information in the form of text, images, videos and audio files from the web and then training algorithms that can produce new information. And since the work of Hollywood screenwriters was already available online, knowledge workers were the first to be affected.
The next step in developing language paradigms is called exact models. They are trained on protected corporate data that is not freely available online. These new paradigms will make it possible to simplify and automate a much larger number of cognitive functions.
For example, the Media Group launched bloomberg Recently an accurate model was created that was trained on all the digital texts produced by Bloomberg employees over the past 40 years. It is expected that this model will make producing news articles for Bloomberg much easier, thus reducing the need for experienced human writers.
Are you an experienced bureaucrat who has written many reports? Journalist who wrote many articles? A communications consultant who has published many press releases? Or a customer employee who has several thousand customer conversations in audio recordings? Then, within a short period of time, your boss will be able to feed this data into an accurate language model that can help your less experienced colleagues do the same job as you.
As with Hollywood screenwriters, in the long run such a development could impair your ability to negotiate wages and safe working conditions. So why should your boss compensate you for your long experience when a low-paid, AI-enabled rookie could do the job, too?
This is nothing new. Since the beginning of the Industrial Revolution, the introduction of new technology and automation in the workplace has brought about dramatic changes for manual laborers.
Over the past 200 years, the introduction of new time-saving technology has generally pushed down wages. This is what the two star economists Daron Acemoglu and Simon Johnson explain in their newly published book “strength and progress». Exceptions to this rule have only arisen where trade unions have a central role in negotiating fair solutions where the gains of new technology also benefit employees.
This is also why the outcome of the strike in Hollywood is so important. The thousands of workers participating in this massive multi-union strike are the front line in the fight against an AI-driven work life and against rising inequality.
We have four tips for anyone who fears AI will automate your job and lower your salary in the long run.
- Engage in dialogue with management and the union In your workplace around the rules of the game to use artificial intelligence. If an accurate language model is trained based on the data generated from your work, it may be appropriate to be compensated for this in terms of salary.
- Use artificial intelligence to improve and simplify your work. Over the summer alone, there have been a number of improvements in AI tools and practical applications within a number of professional fields that you can learn from online. A good starting point is an overarching one List of applications of artificial intelligencerecently launched by the online newsletter Medium.
- Collaboration with colleagues Your goal is to find the best common applications of AI in your business. If you actively share AI experiences with your colleagues, you will more quickly find relevant applications that help you do your work better.
- Also discuss whether the way you use AI for work is safe, ethically sound, and legal. A broad dialogue on these topics is just as important to the management as it is to the employee. At Deloitte, we have developed A framework for the trust-based use of AI in businesses. You, your colleagues and the manager can use it as inspiration.
The spread of generative AI is still at an early stage. We can expect that it will create great changes in working life in the future. Fortunately, in Norway we have a strong tradition of tripartite cooperation based on trust between employees, employers and the state. The latter can help increase productivity and the fair distribution of the gains that this increase brings.
“Web specialist. Lifelong zombie maven. Coffee ninja. Hipster-friendly analyst.”