This provides a set of notes for the Jura Workshop on factchecking. Slides will give extra images and context.
Truth is a sticky concept for many reasons. The Wikipedia page alone has over 10,000 words alone, and this is only a brief summary of philosphy's history on the matter. There are many ways in which statements we would usually consider 'true' in regular conversation can be challenged:
Factchecking is the process of determining whether information is true. An opening question might be: is this important? There have been notable recent exceptions, especially in our political climate. Is it better to believe something because you think it is true, or because you think it serves some other purpose, like making you a happier or better person in some way? Also, in the current culture war should we give up truth as many others have done, and simply say what is politically expedient?
I won't answer these questions here, but I will attempt to provide a summary of my personal working definition of truth.
Disclaimer: I am not a philosopher.
This last point is the most important to factchecking. For the rest of the workshop, we will consider various major sources of information. For each of these, think critically about how you engage with these sources, whether you trust them, and what checks and balances you keep in your mind when sifting through information from them.
The brain is easily tricked by illusions, cognitive biases, anxieties, paranoia and other effects. It's useful to learn a bit of cognitive psychology to understand some of these major effects and begin to notice them. You won't be able to stop these happening, and it takes a fair amount of care to even realise they are happening.
Here is a brief summary of a few worth considering:
The best text for understanding the distortion of truth by news media is 'Manufacturing Consent' by Chomsky and Herman. The main thesis is that propaganda is not necessarily dictated by some top power (although sometimes it is, see Jeff Bezos and the Washington Post recetly). Instead, media which tows the line of the government gets whitelisted and is given insider information they need to perform their service. A useful tool is the 'Overton window', only mentioning ideas that are within the acceptable scope.
Which news media do you engage with? How trustworthy is it? What are you thinking when you're reading articles to make judgements about the quality and truth of the article?
How often do you engage with alternative or independent media? What are the barriers to entry for people getting into this media?
Social media platforms cost an enormous amount of money to run, and almost all of them are free. They make money in three ways: premium features, hosting ads, or selling data. For the latter two to be effective, platforms needs to optimise for engagement over everything else. This is their fundamental goal: making you spend as much time on the platform as possible, which they do with a psychological barrage directly hijacking your decision processes.
Recently this problem has gotten even worse. Nicky Osterweil's article highlights the major steps the tech industry has taken to actively encourage facism for decades. The main points here are the increased use of surveillance capitalism, hypertargeted advertising, AI content, and the encorachment of these technologies into all aspects of our lives.
You should also be mindful of this on a content level. Most influencers on Instagram, Tik Tok and YouTube make money from advertising. If the ads aren't explicit, then they're implicit. Any product with a shown name (and especially a recommendation) are almost certainly advertiser. In my opinion, choose now not to buy products or services advertised in this way.
How can we use social media in beneficial ways? What mental processes do we have running when using social media platforms? Are there better alternatives? Is it time to abandon social media for good?
As a basic summary, generative AI algorithms are fancy averaging operations. ChatGPT trains on a massive amount of data (most of the internet), and when given a prompt, simply outputs the next most common word it would usually see. There are some fancy steps that make the sentences a little more coherent and stop it saying anything problematic. Image tools like Stable Diffusion are slightly more complex but are based on the same fundamental idea.
What's important to remember is that these 'intelligences' don't actually know anything. By design they are incapable of creative thought or novel ideas. They aren't able to factcheck or cross reference. It is very easy to gaslight them into lying or hallucinating.
What is more concerning is the use of AI to doctor or edit footage. Recently, 'The Brutalist' used AI to make the Hungarian voiceover more accurate by shifting the vowels. This AI usage can be very difficult to detect, but there are a number of guides online which give suggestions.
The use of AI has also completely ruined search engines like Google, with the platforms often choosing to do this to themselves. The AI is now also being trained on its own output, completely slopping the waters. A helpful suggestion: use the advanced settings on your search engine to always search for results before 2021 (if applicable).
The scientific method has been a monumental development in history, allowing us to learn many specific facts about the world in the last few hundred years. We would not have developed germ theory, climate science, evolution theory or quantum physics without it. Vaccines work and climate change is real.
However, the process is not infallible and is often subject to misuse. In 1930s Germany, race science was seen as a reasonable and evidence-based discipline with a wide scope of supporting academia. These days, a significant portion of climate science is funded by mining contractors and other carbon-emitting agencies.
Scientific research is also often miscommunicated in the media, with results prone to being sensationalised, placed out of context or simply misunderstood. Any good article reporting on research will link to the paper itself, and the abstract is almost always freely available.
Be extremely wary of individuals claiming to disrupt an entire field, research almost never happens in this way (even with Einstein, despite the common story). Almost remember that scientists in their work will often state results conservatively, and new effects take years of research to be propery included in the literature.