#radiology isn’t really a *clinical* specialty.
it has always been, throughout its history, a #technological specialty.
Thoughts #healthtech #medtwitter #rads ?
Zan Tafakari’s Tweets
This is just as accurate for doctors, lawyers and many others in so-called professional fields.
Nepotism in #medicine (particularly academia) and #healthcare is alive and well.
Quote Tweet
This is just as accurate for novelists, journalists and many others in so-called creative fields twitter.com/Atencio/status…
Show this thread
3/2:
And maybe after the cathartic scapegoating, there will be less chaos.
Show this thread
2/2:
It all depends on how cathartic the crypto space finds his prosecution.
1
Show this thread
5: I'll end by re-iterating.
When you parrot an idea given to you by someone, you're also parroting a part of them.
We've evolved to do that unbelievably well.
1
Show this thread
5: So, truly faithful memetic replication must also include the imitation of the desires that these memes create.
And the imitation of desire is #girard's mimesis.
1
Show this thread
4: Compelling ideas don't just float around in our brains.
We embody them.
They change who are ARE, and how we ACT.
They shape our DESIRES.
1
Show this thread
2: BUT replicating memes is NOT a clean process, as #lukeburgis suggests in his book, #wanting.
He writes that #dawkin's memetics are replicated independent of the source person.
I disagree.
You cannot purely replicate a meme without also somewhat replicating the person.
1
Show this thread
1: Replicating memes requires you to creatively uncover the hidden ideas behind transmitted words.
Deutsch suggests that this is how creativity evolved: to select for humans who could creatively replicate memes (and the important ideas within them) faithfully.
1
Show this thread
i am a stochastic parrot, and so r u
166
262
1,803
In #medicine:
#Biological proof trumps computational proof.
#Computational proof trumps #epidemiological proof.
So don’t just dick around building fancy models.
CONJECTURE THE BIOLOGICAL PROCESS
Some Sunday morning writing.
A book that changed my life before it was written.
↓
24
131
943
These 3 non-negotiables have helped me navigate my swinging emotional seasons and break out of bad habitual cycles.
They give me a sense of intentional stability.
I hope they may help you too.
1
Show this thread
3: REFLECT EVERY DAY
I write into my daily journal. Even if it's just one sentence, I write.
I complete the day, set my intentions for the morning to come, and clear my mind for a restful sleep.
1
Show this thread
2: HAVE A COLD SHOWER EVERY DAY
At the end of every shower, I turn the faucet fully cold for a minute.
I consciously lean into the discomfort, feel the cold, and remind myself of the simple comforts I often take for granted.
1
Show this thread
1: MOVE EVERY DAY
Conscious, not habitual, movement.
Exercise is overrated. Movement is underrated.
1
Show this thread
HAVE SOME NON-NEGOTIABLES
Non-negotiable: a small, deliberate action that you do every. single. day.
No excuses, no missed days, no negotiating with yourself.
They MUST be a little uncomfortable or effortful, and "good for you".
Here are my 3 daily non-negotiables:
1
Show this thread
I often struggled with cycling between runs of good and bad days.
I learnt that I'm influenced by vicious positive/negative habit cycles that drive these actions.
If you're like me, here is 1 concept with 3 examples that helped me break the cycle, and stay level headed. 🧵...
1
Show this thread
I predict a renaissance of sophisticated developments in CLASSICAL #machinelearning tools that will dominate #clinical AI.
... but then again, human beings are not good predictors of an unpredictable world. 6/6
Show this thread
None of those things need generative outputs. None of them NEED #deeplearning.
We just need to derive actionable answers from complex data.
What we need are sophisticated, but EXPLAINABLE models, that are fundamentally based on BIOLOGICAL processes and CLINICAL heuristics. 5/6
1
Show this thread
Clinical medicine is about:
• Diagnostics (often binary: the disease is there or isn't)
• EXPLANATORY risk predictions and pathophysiology (benchtop biological explanations)
• Severity assessments (thin tailed)
• Prognostics (thin tailed)
• EXPLAINABLE management... 4/6
1
Show this thread
It's no surprise that the big leaps in #deeplearning have been in GENERATIVE endeavours
#AlphaFold, #AlphaGo, #GPT4, #Galactica, #ChatGPT, #DALLE2 all CREATE or GENERATE complex outputs.
That's NOT what we need in clinical medicine... 3/6
1
1
2
Show this thread
Deep neural networks are inspired by the architecture of biological neural networks.
Biological neural networks are NOT good predictive machines.
Humans are terrible at explanatory predictions in an unpredictable world.
We're creative, generative, conjectural creatures. 2/6
1
Show this thread
🚨🚨 PSA to everyone in #healthtech or #medtech that's trying to use #deeplearning or #AI for a direct clinical purpose 🚨🚨
... STOP.
And think twice. Heres why... 1/6🧵
1
1
1
Show this thread
4/5:
"Whenever anyone says, 'Life is too short,' I'm like, 'Well, life is the longest thing you know. You don't know anything longer than life.'"
1
2
Show this thread
3/5:
"Conventional thinking is typically right but seldom profitable"
1
Show this thread
2/5:
"The toe that you step on today could be connected to the ass you're going to kiss tomorrow" - anon
1
2
Show this thread
1/5: 3 witty takeaways from the conversation between and Aryeh Bourkoff. Randomly clicked on this episode, blessed that I did. Funny, deep, and one of my favourites:
1
3
7
Show this thread
Key takeaway from the | pod:
Setting and achieving micro tasks (e.g. “clean my desk”) can snowball into bigger actions that may contribute to lifting #depression
TLDR: Huberman explains why JBP’s “clean your room” works.
Practical principle 2:
Add computational thinking to your mental toolbox.
Not everyone needs to be a coder, but being able to think like one helps.
"You talk about productivity hacks. The biggest amplifier, hugest productivity hack is the whole computational language idea"
Show this thread
Practical principle 1: "I did discover, when I was 40 years old, hey you should do some exercise, and that helped me, I think, add a bunch of energy"
Sounds familiar...
Quote Tweet
you experience love by giving it. you experience energy by expending it.
1
Show this thread
Science principle 2:
Simple rules can create complex, emergent phenomena.
These complex phenomena are hard to reverse engineer. Braitenberg's law of "uphill analysis and downhill synthesis" applies. rb.gy/6hw2xl
1
Show this thread
Science principle 1:
The conclusions you make are a function of the data you collect.
"It's easy to answer that question if you have that data"
1
Show this thread
The conversation between and was illuminating.
Here are 2 scientific, and 2 practical takeaways to share from this pod:
1
Show this thread
3: Antifragility in Health
Health is not about "optimising". This is fragile.
It's about thriving in randomness. This is antifragile.
We've evolved within the volatility of mother nature.
Don't optimise for stability. Thrive in volatility. #antifragile #nnt #nassimtaleb
Show this thread
2: Creative Knowledge in Science.
Science isn't about "uncovering" knowledge about the real world.
It's about creating, then testing out our creative conjectures.
Real science creates explanations of the world. It's not epidemiological or meta-analytic. #daviddeutsch
1
Show this thread