Raghav Agrawal @impactology@mastodon.social
@impactology.bsky.social
Former Diamond Dealer | Independently explored research in design, education, and cognitive science; now bringing those ideas to life via UX
https://twitter.com/impactology
https://twitter.com/impactology
Rowboat
A lightning fast tool for understanding large datasets
rowboat.net/product/
"Open one million rows of data in one second. Your data stays private in any browser or on your desktop. No code or complex queries are needed to visualize your data"
A lightning fast tool for understanding large datasets
rowboat.net/product/
"Open one million rows of data in one second. Your data stays private in any browser or on your desktop. No code or complex queries are needed to visualize your data"
November 21, 2025 at 11:20 AM
Rowboat
A lightning fast tool for understanding large datasets
rowboat.net/product/
"Open one million rows of data in one second. Your data stays private in any browser or on your desktop. No code or complex queries are needed to visualize your data"
A lightning fast tool for understanding large datasets
rowboat.net/product/
"Open one million rows of data in one second. Your data stays private in any browser or on your desktop. No code or complex queries are needed to visualize your data"
Reposted by Raghav Agrawal @impactology@mastodon.social
Hi everyone! ✨
I'm an independent artist & designer who enjoys creating scientifically-informed & fantastical visuals in 2D & 3D.
Prev. the establishing Art Director of the award-winning Quanta Magazine.
OlenaShmahalo.com
Stuff I do ↓
#SciArt #Illustration #3Dart #Design
🧪 ⚛️ 🔭 🐡
I'm an independent artist & designer who enjoys creating scientifically-informed & fantastical visuals in 2D & 3D.
Prev. the establishing Art Director of the award-winning Quanta Magazine.
OlenaShmahalo.com
Stuff I do ↓
#SciArt #Illustration #3Dart #Design
🧪 ⚛️ 🔭 🐡
November 15, 2024 at 3:10 PM
Hi everyone! ✨
I'm an independent artist & designer who enjoys creating scientifically-informed & fantastical visuals in 2D & 3D.
Prev. the establishing Art Director of the award-winning Quanta Magazine.
OlenaShmahalo.com
Stuff I do ↓
#SciArt #Illustration #3Dart #Design
🧪 ⚛️ 🔭 🐡
I'm an independent artist & designer who enjoys creating scientifically-informed & fantastical visuals in 2D & 3D.
Prev. the establishing Art Director of the award-winning Quanta Magazine.
OlenaShmahalo.com
Stuff I do ↓
#SciArt #Illustration #3Dart #Design
🧪 ⚛️ 🔭 🐡
Tl; DR
Understand discrete nature of tokenization & statistical limits behind modeling continuous phenomena
Understand the living evolving context that static training data can't capture
Make creative leaps & connections between unexpected disparate things that statistical models can't generate
Understand discrete nature of tokenization & statistical limits behind modeling continuous phenomena
Understand the living evolving context that static training data can't capture
Make creative leaps & connections between unexpected disparate things that statistical models can't generate
What cannot be statistically modelled in an LLM’S vector space?
What is unquantifiable in your work for which you cannot create correlations and patterns? What kind of work cannot be reduced to probability distributions over tokens
What is unquantifiable in your work for which you cannot create correlations and patterns? What kind of work cannot be reduced to probability distributions over tokens
November 20, 2025 at 5:27 PM
Tl; DR
Understand discrete nature of tokenization & statistical limits behind modeling continuous phenomena
Understand the living evolving context that static training data can't capture
Make creative leaps & connections between unexpected disparate things that statistical models can't generate
Understand discrete nature of tokenization & statistical limits behind modeling continuous phenomena
Understand the living evolving context that static training data can't capture
Make creative leaps & connections between unexpected disparate things that statistical models can't generate
What cannot be statistically modelled in an LLM’S vector space?
What is unquantifiable in your work for which you cannot create correlations and patterns? What kind of work cannot be reduced to probability distributions over tokens
What is unquantifiable in your work for which you cannot create correlations and patterns? What kind of work cannot be reduced to probability distributions over tokens
November 19, 2025 at 4:50 PM
What cannot be statistically modelled in an LLM’S vector space?
What is unquantifiable in your work for which you cannot create correlations and patterns? What kind of work cannot be reduced to probability distributions over tokens
What is unquantifiable in your work for which you cannot create correlations and patterns? What kind of work cannot be reduced to probability distributions over tokens
Reposted by Raghav Agrawal @impactology@mastodon.social
I don't think this science communication resource gets enough love.
OpenMoji offers 4,000+ free, #opensource emojis (CC BY-SA 4.0), with categories for healthcare, climate, UI...
Challenge: find Greta Thunberg and a Viennese coffee house. ☕
🔗 openmoji.org #SciComm #Design
OpenMoji offers 4,000+ free, #opensource emojis (CC BY-SA 4.0), with categories for healthcare, climate, UI...
Challenge: find Greta Thunberg and a Viennese coffee house. ☕
🔗 openmoji.org #SciComm #Design
November 18, 2025 at 10:41 AM
I don't think this science communication resource gets enough love.
OpenMoji offers 4,000+ free, #opensource emojis (CC BY-SA 4.0), with categories for healthcare, climate, UI...
Challenge: find Greta Thunberg and a Viennese coffee house. ☕
🔗 openmoji.org #SciComm #Design
OpenMoji offers 4,000+ free, #opensource emojis (CC BY-SA 4.0), with categories for healthcare, climate, UI...
Challenge: find Greta Thunberg and a Viennese coffee house. ☕
🔗 openmoji.org #SciComm #Design
Regenerative UX by @alexgrantwright.bsky.social
uxdesign.cc/regenerative...
"The methods commonly in use today have remained remarkably stagnant for the past 10–15 years: qualitative user interviews, journey maps, personas, prototypes, usability tests, diary studies, and so forth"
uxdesign.cc/regenerative...
"The methods commonly in use today have remained remarkably stagnant for the past 10–15 years: qualitative user interviews, journey maps, personas, prototypes, usability tests, diary studies, and so forth"
Regenerative UX
“Nothing is easier than to formulate high ideals, but few things are more difficult than to discover the means whereby those ideals can be…
uxdesign.cc
November 14, 2025 at 8:18 AM
Regenerative UX by @alexgrantwright.bsky.social
uxdesign.cc/regenerative...
"The methods commonly in use today have remained remarkably stagnant for the past 10–15 years: qualitative user interviews, journey maps, personas, prototypes, usability tests, diary studies, and so forth"
uxdesign.cc/regenerative...
"The methods commonly in use today have remained remarkably stagnant for the past 10–15 years: qualitative user interviews, journey maps, personas, prototypes, usability tests, diary studies, and so forth"
Badass : Making Users Awesome by Kathy Sierra
sobrief.com/books/badass
Badass: Making Users Awesome by Kathy Sierra at Mind the Product San Francisco
www.youtube.com/watch?v=iBtc...
"Instead of solely improving your product, concentrate on making your users better at using it"
sobrief.com/books/badass
Badass: Making Users Awesome by Kathy Sierra at Mind the Product San Francisco
www.youtube.com/watch?v=iBtc...
"Instead of solely improving your product, concentrate on making your users better at using it"
November 12, 2025 at 7:18 PM
Badass : Making Users Awesome by Kathy Sierra
sobrief.com/books/badass
Badass: Making Users Awesome by Kathy Sierra at Mind the Product San Francisco
www.youtube.com/watch?v=iBtc...
"Instead of solely improving your product, concentrate on making your users better at using it"
sobrief.com/books/badass
Badass: Making Users Awesome by Kathy Sierra at Mind the Product San Francisco
www.youtube.com/watch?v=iBtc...
"Instead of solely improving your product, concentrate on making your users better at using it"
Although I will say this, one time summative assessments have zero resemblance to knowledge and skill usage in real world use.
Think about this, if you get any error in your daily work do you just scrap all of it completely or just fix that particular thing and iterate?
Think about this, if you get any error in your daily work do you just scrap all of it completely or just fix that particular thing and iterate?
Got radicalized by teachers teaching to the test for memorization based exams. Got me researching about project based learning, open book assessments, open education
Got deradicalized again by realizing those exams are there for a reason : easiest to scale for millions
Got deradicalized again by realizing those exams are there for a reason : easiest to scale for millions
“what radicalized you” idk paying attention
November 10, 2025 at 2:30 PM
Although I will say this, one time summative assessments have zero resemblance to knowledge and skill usage in real world use.
Think about this, if you get any error in your daily work do you just scrap all of it completely or just fix that particular thing and iterate?
Think about this, if you get any error in your daily work do you just scrap all of it completely or just fix that particular thing and iterate?
If Vannevar Bush, J.C.R. Licklider, Douglas Engelbart, Ted Nelson, and Alan Kay were to work at modern Series A startups as designers, what would their workflow have been like?
November 10, 2025 at 11:53 AM
If Vannevar Bush, J.C.R. Licklider, Douglas Engelbart, Ted Nelson, and Alan Kay were to work at modern Series A startups as designers, what would their workflow have been like?
What kind of software business makes scaling unnecessary?
What kind of software output generates so much value density per user, per employee that the company doesn’t need to scale headcount, bureaucracy, or user base to sustain itself?
What kind of software output generates so much value density per user, per employee that the company doesn’t need to scale headcount, bureaucracy, or user base to sustain itself?
November 9, 2025 at 11:23 AM
What kind of software business makes scaling unnecessary?
What kind of software output generates so much value density per user, per employee that the company doesn’t need to scale headcount, bureaucracy, or user base to sustain itself?
What kind of software output generates so much value density per user, per employee that the company doesn’t need to scale headcount, bureaucracy, or user base to sustain itself?
Got radicalized by teachers teaching to the test for memorization based exams. Got me researching about project based learning, open book assessments, open education
Got deradicalized again by realizing those exams are there for a reason : easiest to scale for millions
Got deradicalized again by realizing those exams are there for a reason : easiest to scale for millions
“what radicalized you” idk paying attention
November 9, 2025 at 10:50 AM
Got radicalized by teachers teaching to the test for memorization based exams. Got me researching about project based learning, open book assessments, open education
Got deradicalized again by realizing those exams are there for a reason : easiest to scale for millions
Got deradicalized again by realizing those exams are there for a reason : easiest to scale for millions
Reposted by Raghav Agrawal @impactology@mastodon.social
I have previously worked in an architecture team and it was great. AND...
Software Design Is NOT Software Architecture (& That’s A Good Thing!) | @emilybache.com
📽️ AVAILABLE NOW
Watch HERE ➡️ youtu.be/idGM8fvtT8U
📽️ AVAILABLE NOW
Watch HERE ➡️ youtu.be/idGM8fvtT8U
Software Design Is NOT Software Architecture (& That’s A Good Thing!)
YouTube video by Modern Software Engineering
youtu.be
October 23, 2025 at 8:47 AM
I have previously worked in an architecture team and it was great. AND...
Reposted by Raghav Agrawal @impactology@mastodon.social
AI-assisted coding is like the blind men and the elephant. Everyone's experiencing it from different vantage points.
Makes it genuinely hard to know whose experience is worth learning from.
Sketched a graph to illustrate & wrote about the patterns I'm seeing.
blog.robbowley.net/2025/10/26/y...
Makes it genuinely hard to know whose experience is worth learning from.
Sketched a graph to illustrate & wrote about the patterns I'm seeing.
blog.robbowley.net/2025/10/26/y...
October 28, 2025 at 8:03 AM
AI-assisted coding is like the blind men and the elephant. Everyone's experiencing it from different vantage points.
Makes it genuinely hard to know whose experience is worth learning from.
Sketched a graph to illustrate & wrote about the patterns I'm seeing.
blog.robbowley.net/2025/10/26/y...
Makes it genuinely hard to know whose experience is worth learning from.
Sketched a graph to illustrate & wrote about the patterns I'm seeing.
blog.robbowley.net/2025/10/26/y...
Reposted by Raghav Agrawal @impactology@mastodon.social
Attaching a code-generating firehose to our development process that was largely trained on short-form code examples means we need to really stay on top of the bigger picture as it emerges one test and one refactoring at a time.
codemanship.wordpress.com/2025/10/29/t...
codemanship.wordpress.com/2025/10/29/t...
The AI-Ready Software Developer #14 – Continuous Architecture
They say a journey of a thousand miles starts with a single step, but we still need to look where we’re going. One complaint that’s often leveled at micro-iterative development practice…
codemanship.wordpress.com
October 29, 2025 at 8:53 AM
Attaching a code-generating firehose to our development process that was largely trained on short-form code examples means we need to really stay on top of the bigger picture as it emerges one test and one refactoring at a time.
codemanship.wordpress.com/2025/10/29/t...
codemanship.wordpress.com/2025/10/29/t...
Reposted by Raghav Agrawal @impactology@mastodon.social
I wrote a blog post today "The Microskills Theory of Programming": coding-is-like-cooking.info/2025/10/the-...
The Microskills Theory of Programming - Coding Is Like Cooking
How to Succeed with Technical Coaching Software Developers are constantly shifting between activities. It might all be ‘coding’ but when you look more closely there are hundreds of distinct activities...
coding-is-like-cooking.info
October 31, 2025 at 12:08 PM
I wrote a blog post today "The Microskills Theory of Programming": coding-is-like-cooking.info/2025/10/the-...
I found the term for what I've been doing tacitly for years: conceptual blending. It's the practice of intentionally mixing disparate disciplines to discover novel intersections that don't exist in any single field.
en.wikipedia.org/wiki/Concept...
en.wikipedia.org/wiki/Concept...
November 1, 2025 at 6:57 PM
I found the term for what I've been doing tacitly for years: conceptual blending. It's the practice of intentionally mixing disparate disciplines to discover novel intersections that don't exist in any single field.
en.wikipedia.org/wiki/Concept...
en.wikipedia.org/wiki/Concept...
Transferable Skills Are Cool. But Do They Transfer? – Education Rickshaw educationrickshaw.com/2020/03/27/t...
October 30, 2025 at 5:30 PM
Transferable Skills Are Cool. But Do They Transfer? – Education Rickshaw educationrickshaw.com/2020/03/27/t...
A cognitive science research question
At what level of performance of a skill can we identify pattern of thinking that can be translated from one domain to another
A question of transferrability of a skill
At what level of performance of a skill can we identify pattern of thinking that can be translated from one domain to another
A question of transferrability of a skill
October 30, 2025 at 5:25 PM
A cognitive science research question
At what level of performance of a skill can we identify pattern of thinking that can be translated from one domain to another
A question of transferrability of a skill
At what level of performance of a skill can we identify pattern of thinking that can be translated from one domain to another
A question of transferrability of a skill
October 28, 2025 at 8:20 AM
In your field, what’s the smallest valuable act you repeat that actually makes you better? If you had to turn that act into a daily exercise for others, what would it look like?
Purpose of an exercise or an assignment is to compress the maximum amount of retention of concepts principles in a single exercise such that there is a proper gestalt of understanding that a person can form, that when repeated it forms an automatic skill, that the mind retains
October 27, 2025 at 2:17 PM
In your field, what’s the smallest valuable act you repeat that actually makes you better? If you had to turn that act into a daily exercise for others, what would it look like?
One thing I realized, difference between quality of an ivy professor's assignments vs online course instructors
October 27, 2025 at 2:12 PM
One thing I realized, difference between quality of an ivy professor's assignments vs online course instructors
Reposted by Raghav Agrawal @impactology@mastodon.social
'This report looks at who gets to be an artist in the UK today, and considers whether access to a career in this vital industry is equitable and accessible'.
Such a relief we've got the White Paper on post-16 skills and education deeply committed to rectifying this anomaly in England... 2/2
Such a relief we've got the White Paper on post-16 skills and education deeply committed to rectifying this anomaly in England... 2/2
Who gets to be an artist? | Artquest
This report looks at who gets to be an artist in the UK today, and considers whether access to a career in this vital industry is equitable and accessible Student working in the painting studio, 2022,...
artquest.org.uk
October 27, 2025 at 12:09 PM
'This report looks at who gets to be an artist in the UK today, and considers whether access to a career in this vital industry is equitable and accessible'.
Such a relief we've got the White Paper on post-16 skills and education deeply committed to rectifying this anomaly in England... 2/2
Such a relief we've got the White Paper on post-16 skills and education deeply committed to rectifying this anomaly in England... 2/2
Reposted by Raghav Agrawal @impactology@mastodon.social
'Are you from a professional family, went to private school, have a postgraduate degree, and live in London with a load of cash in the bank?
If so, congrats – you’re three times more likely to become an artist than anyone else.' 1/2
If so, congrats – you’re three times more likely to become an artist than anyone else.' 1/2
Who gets to be an artist?
Are you from a professional family, went to private school, have a postgraduate degree, and live in London with a load of cash in the bank?
wonkhe.com
October 27, 2025 at 12:09 PM
'Are you from a professional family, went to private school, have a postgraduate degree, and live in London with a load of cash in the bank?
If so, congrats – you’re three times more likely to become an artist than anyone else.' 1/2
If so, congrats – you’re three times more likely to become an artist than anyone else.' 1/2
What customer segments consistently reject products designed by elite-educated teams because the solutions miss their actual pain points?
October 26, 2025 at 6:29 PM
What customer segments consistently reject products designed by elite-educated teams because the solutions miss their actual pain points?
KG-PLPPM: A Knowledge Graph-Based Personal Learning Path Planning Method Used in Online Learning
mdpi.com/2079-9292/14...
Using knowledge graphs, the method personalizes learning paths by mapping concept similarity, learner mastery, and planning optimal routes.
mdpi.com/2079-9292/14...
Using knowledge graphs, the method personalizes learning paths by mapping concept similarity, learner mastery, and planning optimal routes.
MOOCCube: A Large-scale Data Repository for NLP Applications in MOOCs
aclanthology.org/2020.acl-mai...
"A large-scale data repository of over 700 MOOC courses, 100k concepts, 8 million student behaviors with an external resource"
aclanthology.org/2020.acl-mai...
"A large-scale data repository of over 700 MOOC courses, 100k concepts, 8 million student behaviors with an external resource"
October 25, 2025 at 7:37 AM
KG-PLPPM: A Knowledge Graph-Based Personal Learning Path Planning Method Used in Online Learning
mdpi.com/2079-9292/14...
Using knowledge graphs, the method personalizes learning paths by mapping concept similarity, learner mastery, and planning optimal routes.
mdpi.com/2079-9292/14...
Using knowledge graphs, the method personalizes learning paths by mapping concept similarity, learner mastery, and planning optimal routes.