Ah yes, the known unknowns.

The discussion reminds me of a talk Zizek gave in which he discusses the speech Rumsfeld gave regarding the evidence Iraq supplying weapons to terrorist[0].

Zezik argues the unknown knowns are far more interesting (and the reason why USA was losing in Iraq). While Rumsfeld focused on the unknown unknowns.

I've noticed that domain experts who implicitly know the the known unknowns of their field distrust LLMs because they can identify their shortcomings. Those subtle mistakes LLMs make. I argue this is why domain experts using LLMs get such a boost. They can identify and avoid pitfalls sometimes before they happen. But in other fields the same people are in awe of LLM capabilities precisely because the known unknowns are a mystery.

The Unknown Unknowns of LLMs are the IMO the most interesting. The so called emergent capabilities of the technology. The use of LLMs in others fields such as biology, eg in protein language models, is really cool.

Everyone focuses on replacement of people workers when I think opening new fields of work for humans should be the goal of LLMs by leveraging the tech to discover.

The other interesting caregory is unknown knows. But that's another topic for another time.

[0] https://en.wikipedia.org/wiki/There_are_unknown_unknowns

Comments URL: https://news.ycombinator.com/item?id=48452678

Points: 3

# Comments: 0

While the definition changes, the expertise shifts and with it the field. Computers eventually became statisticians and data scientists. Printers became graphic designers.

What I found most interesting is that when positions undergo such evolution (printer -> graphic designer), a number of skills which were previously different expertise altogether, combine to create a new field. In other words, a new multidisciplinary field is born.

I think a good example is data science, the field at it's core is applied statistics using modern techniques such as data management and computing [0].

The question is, what is the new evolution of a programmer? Lots of folks like to use the term "engineer", and previously I thought this was silly. But now with LLMs, maybe that is a good descriptor; software engineer.

[0] https://www.welcometothejungle.com/en/articles/story-origin-...

https://www.nature.com/articles/s41541-024-00950-9

Its bizarre and even frustrating to see basic bioinformatics methodology referred to as AI.

AI had become a catch all term for... Everything.

Reading it now, great read!

https://news.ycombinator.com/item?id=41438060

That's the answer IMO, yeah now there's two UX to maintain but it's a step forward.

Would therapy work on an LLM?

True automation of scientific research requires true AI.

I'm not really sure why tech people keep suggesting scientists aren't incorporating the latest tech.

We are all about automation. The issue is funding.

Eg here's a nice paper discussing Bayesian phylo combined with rexord sampling. https://www.cambridge.org/core/journals/paleobiology/article...

I'm in an adjacent field so take it with a grain of salt.

Testing relies on amplifying many parts of the human DNA not just a small fragment.

The smaller the fragment the less precise the comparison.

So no, floating DNA does not meaningfully impact testing. It can impact PCR assays that target small regions for identification eg of viruses.

Here's an old forum post discussing it

https://tl.net/forum/brood-war/352588-my-replays-are-broken

Here's a hacky patch that also documents some of the possible issues

https://liquipedia.net/starcraft/Replay_Fix_for_1.16.1

It seems like the bug is related to pausing the game.

https://news.ycombinator.com/item?id=21920508