'Evaluating evidence in the context of uncertainty is difficult. Communicating such evidence is more difficult yet.' -- Donald Berry

ok we're actually getting to the point where we can turn piles of unstructured data into insights you can see 🗺️ @[email protected] and I have been building out this technique and now we can't stop throwing CSV files into it 🪄 let's take a look at 765 free response survey answers 📍

Writing assistants help you write predictable prose. This is true by definition: the software predicts what you want to say. This could be helpful, but it also encourages mediocre, cliché writing.

Late 90s/early 2000s when I was a student: 'media studies' was the go to whipping boy/punchbag for people mocking 'useless degrees' 2016 onwards: Woah! If only the population were more savvy about how the media works and affects us, we wouldn't be in all these messes

"Phew! That's a nasty leak. Thank goodness it's not at our end of the boat". An image that's as powerful in the age of the climate crisis as it was in the Great Depression. Then, as now, "it's not our problem" was a spectacular misjudgement. Thanks to Cllr David Ward 🌹 for posting

Some words that might usefully be remembered today, on why Opposition matters - and why critics are not "enemies of the people", to be bundled up with "criminal gangs" and "people traffickers". From A.L. Lowell's 1908 text, "The Government of England".

It’s a common counterpoint: “But we can’t let China get ahead.” On #YourUndividedAttention, assistant professor Jeffrey Ding and journalist Karen Hao explain the realities of Chinese #AI development and assess the stakes for the multi-national race. bit.ly/3OZTK9W

Models, like men, sometimes do the right thing for the wrong reason.

“The data may not contain the answer. The combination of some data and an aching desire for an answer does not ensure that a reasonable answer can be extracted from a given body of data.” — John Tukey

where is the other face???

“You can construct any binary circuit using nothing but floating point subtraction.” orlp.net/blog/subtracti…

The "Ship of Theseus" article has been edited 1792 times since it was created in July of 2003. At present, 0% of the phrases in the original article (seen below) remain.

We're excited to announce the Inter-Species Phonetic Alphabet. An AI-based, precise, concise, and interpretable system designed to transcribe animal sounds into text, this opens myriad new possibilities in bioacoustics. arxiv.org/abs/2402.03269 github.com/earthspecies/i…"

Are medical studies being written with ChatGPT? Well, we all know ChatGPT overuses the word "delve". Look below at how often the word 'delve' is used in papers on PubMed (2023 was the first full year of ChatGPT).

Psychedelic cryptography, or "hiding information in images that could only be understood when tripping": qri.org/blog/psycrypto… via Scope of Work, cc Michael Pollan

The more deprived the constituency, the lower the turnout. Only one truly effective solution to this - which lots of people will hate ('The answer is to enthuse people, not compel them') and so won't ever happen - COMPULSORY VOTING. Source: Tom Calver thetimes.com/uk/politics/ar…

From Deep Learning with Python: - Deep learning research is an evolution process driven by poorly-understood empirical results - Math in DL papers is usually worthless and was placed there purely as a signal of seriousness - The key to good research is understanding what

The way to predict the future is to look at the past not as a source of reference situations, but as a source of meta-modeling insights. The future won't be like the past, so you can't just assume that a similar situation will play out similarly. But it is by backtesting on past

A surprising & little-known results in classical statistics: mean (μ) and median (m) are within one std deviation: |μ−m| ≤ σ For unimodal densities, bound is even tighter |μ−m| ≤ 0.7746 σ This beautiful results first appeared in a 1932 paper by Hotelling & Solomons 1/3

'We can only connect the dots that we collect.' -- Amanda Palmer