«We need source criticism more then ever now.»

In the season finale of MetaDAMA, we dive deep into the intersection of philosophy, history, and artificial intelligence with guest Sune Selsbæk-Reitz, tech philosopher with a background in both history and philosophy.

Sune introduces the provocative concept of “Promptism”, which is our era’s version of positivism, where we believe that truth can be extracted from language models simply by phrasing the question correctly. But just as historians have learned through centuries of source criticism, we must ask the critical questions: Who trained this model? On what data? With what biases?

Here are Winfrieds key takeaways:

• Are numbers and data points neutral? Or can they be used to convey a message, or even a certain philosophical view point?
• Philosophy is important in data. Here is an example:
  
  • Data Governance according to Immanuel Kant - best possible governance
  • Data Governance according to Utilitarism - focus on business value
• Lessons from history studies: question the authenticity of your sources. Who wrote it? For what purpose? Why are you reading it? - same lessons apply to data.
• That is why we need principles and values in AI ethics.
• Over-reliance on the objectivity of math - is math binary? Right or wrong? - this has been introduced into algorithmic thinking and AI.
• This is the reason why «algorithmic authority» is an issue - because the algorithm says «right» doesn’t mean it is right.
• Our mindset is constantly evolving. That’s why we cannot predict tomorrow’s bias. We need to ensure that our systems evolve with us.

What is the real purpose of AI systems? Are the core values only efficiency, automation, or is it human dignity or autonomy?

Promptism:

• A new way of «positivism: Just because its written down its true»
• Promptism is my term for a subtle but growing mindset around the globe, that you can extract truths from a language model, just by wording your prompt well.»
• LLMs are very fluent and flattering - they say what people want to hear.
• «That’s what Large Language models are: You are not getting the truth. You are just getting the most common answer.»
• Objectivity is a myth. It is always subjective, so we need to read not only the text but also peoples intentions with the text.
• Responsibility for understanding at the limitations and needed criticism of LLM output is shared.
• Producers have a responsibility to ensure that you can know, when models are hallucination, guardrails in models to ensure that output is not looked at as the truth.
• Readers are responsible to learn how to read and understand machines.
• Consumers need rot push for transparency.
• Without accountability, trust is eroding.
• Politeness is a way for machines to ensure that users keep using them.
• Shouldn’t rather an LLM as a «conversation partner» challenge you? Disagreement is part of learning.
• «Agreeableness is addictive.»
• People are starting to get influenced by how LLMs are writing. It changes written conversation.
• Is language narrowed down to a certain path defined through AI? Is language becoming controllable?
• LLMs affect our lives in the way we read, write, talk, even think.
• There is a worldview baked into the system.
• Literacy means also critical thinking.

«Hvis folk tror at data faget er forskjellig mellom industrier så er det ikke det. Data er Data. Forretningskunnskapen er forskjellig. / If people think data work is different across industries, it isn’t. Data is data. Business knowledge is what changes.»

In this episode of MetaDAMA, we dive into the data strategy of Statens Vegvesen with Bjørn Broum. With 30 years of experience in the data field, Bjørn shares valuable insights.

Bjørn takes us on the journey from Statens Vegvesen’s first data strategy to today’s revised version and explains why a strategic approach to data is essential for success. The conversation also touches on critical questions around privacy, skills development, and organizational challenges. Bjørn shares his most important lesson: keep technical complexity low in the early stages, focus on demonstrating value, and remember that decisions are made with or without data, so make data available when it’s needed!

Here are some key takeaways:

• We need a mindset shift in organizations. Think about the following:
  • You need to keep focus on operational success, that is what you are measured by.
  • You need to understand how you can impact your architecture, before the actual impact happens.
  • New services needs to be build around data. Data cannot be an afterthought,
• You need to work actively with your executive management to ensure that data becomes a stable, not something that is hype-driven.
• Your data infrastructure is what you need consistent focus and buy-in on. This is the foundation that can support the «hype» on top of it.

Statens Veivesen

• Long standing tradition dating back to 1864, tasked to build and maintain road infrastructure in Norway.
• Maintaining 10.500 km of road infrastructure, including about 6000 bridges and 600 tunnels.
• Utilization of sensors to predict risks and impact of e.g. rockfall, snowdrift, avalanches that all could impact traffic and road security.
• Research on how Norway can prepare for autonomous vehicles and traffic safety.
• Modern cars can be used as sensors. You can e.g. get information on how many people are in a car.
• There is a focus on ensuring personal data protection while maximizing insights from data.
• Ensure that the organization develops in a direction that can address and handle new challenges (data-driven).
• Remember that decisions are made with or without data. If you want data-informed decision data needs to be available for decision making, when needed.

Data Strategy

• It’s a living strategy, that needs to stay updated.
• Data Strategy is always designed to accompany for a need, and that need can change over time.
• Data Strategy needs to hold a big-picture that people can relate their work to, that is recognizable.
• Digital value chain needs to be understood also towards the cross-functional needs across the organization.
• Conways law needs to be recognized.
• The value of data is realized in business. Therefore Data Strategy is build on business objectives.
• Be deliberate with your strategy: what do you need to focus on? For veivesenet this was e.g. to establish a common language and vocabulary.
• Statens Veivesen also makes data accessible for research. This needs to be reflected.
• AI is a topic as well, including data readiness for AI.
• Competency and skills are an important part of data strategy.
• Don’t focus on dat roles, but rather upskilling in your organization.
• Understanding boundaries, goals, and strategic priorities is key when working with data strategy.
• Scaling and dimensions need to be strategically understood.
• Focus on the value of data and the business value it potentially generates, not technology.

Bjørns' blog

«A lot of things break with scale.»

In our latest conversation with Mikkel Dengsøe, co-founder of SYNQ and former Head of Data, Ops and Financial Crimes at Monzo Bank, we explore the secrets behind effective scaling of data teams.

Mikkel reveals surprising statistics based on his analysis of over 10,000 LinkedIn data points and valuable insights from Monzo’s scaling journey, where the data team grew from 30 to over 100 people in just two years.

We discuss the critical balance between central data teams and domain experts, the importance of career paths for individual contributors (not just managers), and how data professionals can succeed by building relationships with stakeholders who involve them early in strategic processes.

Here are our key takeaways:

Data Teams

• There are some high-level questions you need to ask yourself when building, structuring or scaling a new data team
• This includes how big the team should be, also relatively too your organizations size and other teams, how it should be composed and structured, etc.
• A good idea is to collect data to create a benchmark.
• Benchmarks can be hard to combine and are a moving target, but they are nevertheless valueable.
• Most importantly, you need to ask yourself: WHY do we need to scale our data team?
• Involve people actively in setting the goals based on your WHY.
• Mikkel collected over 10.000 data points from companies on Linkedin. Here’s what he found:
  
  • Median % of data people in companies out of overall staff is 1-4%.
  • Data team relative to engineering team varies between 1 data person per 10 engineers to 1 in 3.
  • From the benchmark it is evident that data governance roles only appear in lager companies.
  • In marketplace companies the effect of data on the business value is easiest to track. Therefore they seem more willing to invest In data teams.
• Investment in data means investment in your business. The consequences of not investing in data will be tangible in your business.
• Find a risk based approach to data as well. At what level can you balance investment, outcome and risk?
• Be cautious of «pseudo-data teams» - teams in a Business unit that do kind-of data work, but are not aligned with the organization.
• Be clear on the skills and competencies you need. What is a data analyst? What does a data scientist do in your organization?
• It is important to have a clear and consistent internal career ladder. Make it visible and understandable what is expected from each role on your team and don’t change these expectations too often.
• Create pulse checks to understand what people are happy about and what not.

Scaling Data Teams

• «Golden Nugget Awards» to showcase good data work every month. These were added to a database, so every new employee could evaluate them to see what good looks like.
• Write down your progression framework to get clear about your ideas and how people excel in your organization.
• You can show open what work lead to promotions. That can be engaging for people to follow in these tracks.
• Hub-n-Spoke model, where people rotate in and out of the central team and the distributed teams.
• Citizen developer programs are a way for larger organizations to scale data work. But It bears risk related to data literacy.
• Don’t try to enable everyone, but enable those that are motivated.
• «You shouldn’t necessarily force people into management to progress.»
• Senior technical careers can ensure an advanced level of quality. Which is a different way of scaling your data team.
• You need a career ladder for professionals that is independent from management careers.
• Create rituals that make good work stand out.