Christian Gustavsson
@christiangu.bsky.social
AI, Cybersecurity, and Defense. PhD Student at Linköping University. Loving long runs and even longer bike rides.
💡 In 1960, J.C.R. Licklider published Man-Computer Symbiosis, imagining humans and computers as partners. Sixty-five years later, his vision looks prophetic, outlining the existence of generative AI, cloud, GUIs, search engines, and voice assistants.
The 1960 Paper That Saw the Future of AI
I stumbled upon the most remarkable paper forecasting our present-day emerging relationship with generative AI. To kick off this autumn, I really want to spread this fundamental read on AI.
phdbytes.substack.com
August 18, 2025 at 8:39 AM
💡 In 1960, J.C.R. Licklider published Man-Computer Symbiosis, imagining humans and computers as partners. Sixty-five years later, his vision looks prophetic, outlining the existence of generative AI, cloud, GUIs, search engines, and voice assistants.
What I’m reading this summer - and why! Ping @warmatters.bsky.social
Reading list for this summer!
What I’m reading this summer - and why. Most in English, some in Swedish. Some fiction, some non-fiction. I hope you find something that intrigues you!
phdbytes.substack.com
July 12, 2025 at 9:26 AM
What I’m reading this summer - and why! Ping @warmatters.bsky.social
När politiken ger sig på barns liv för att vinna poänger, då har en farlig gräns passerats. Om det skriver jag i en ny ledarkrönika i Corren.
CHRISTIAN GUSTAVSSON: Politikens nya passerade gräns: Att ge sig på någons barn
När barn blir måltavlor för att komma åt politiska motståndare så har en gräns passerats. Finns det ens en väg tillbaka?
www.corren.se
July 11, 2025 at 10:51 AM
När politiken ger sig på barns liv för att vinna poänger, då har en farlig gräns passerats. Om det skriver jag i en ny ledarkrönika i Corren.
Technology isn’t enough. Without people, security fails. New post on the human firewall and security culture.
Why Technology Alone Won't Save Us: The Human Firewall and Security Culture
The first post in a two part-series on why humans are the weakest link in cybersecurity, and how they could be the greatest asset.
phdbytes.substack.com
June 26, 2025 at 7:55 AM
Technology isn’t enough. Without people, security fails. New post on the human firewall and security culture.
Are we held back by our need to be liked? The Courage to Be Disliked says yes — and dares us to choose freedom instead.
The Courage To Be Disliked
Ichiro Kishimi and Fumitake Koga, 2013
phdbytes.substack.com
June 22, 2025 at 4:48 PM
Are we held back by our need to be liked? The Courage to Be Disliked says yes — and dares us to choose freedom instead.
💻 Can tiny, local LLMs help you code in air-gapped environments? Turns out - yes. No cloud, no leaks, just offline AI magic.
Can Small Language Models Really Code?
AI-assistance in coding - without relying on the internet?
phdbytes.substack.com
June 18, 2025 at 4:56 AM
💻 Can tiny, local LLMs help you code in air-gapped environments? Turns out - yes. No cloud, no leaks, just offline AI magic.
There’s a book that’s nearly 90 years old, and still offers some of the best advice available today on how to be human - and how to interact with others. No LinkedIn clichés. No AI. No lifehacks. No 10X Your Network!
How to Win Friends and Influence People
Dale Carnegie, 1936
phdbytes.substack.com
June 15, 2025 at 9:44 AM
There’s a book that’s nearly 90 years old, and still offers some of the best advice available today on how to be human - and how to interact with others. No LinkedIn clichés. No AI. No lifehacks. No 10X Your Network!
In the world of artificial intelligence, bigger has often meant better. Are we on the brink of chasing the complete opposite, realizing that less could actually be more?
Smaller, Smarter, Safer: Why the Future of Healthcare AI Might Be Small
In the world of artificial intelligence, bigger has often meant better. Are we on the brink of chasing the complete opposite, realizing that less could actually be more?
phdbytes.substack.com
June 11, 2025 at 3:38 PM
In the world of artificial intelligence, bigger has often meant better. Are we on the brink of chasing the complete opposite, realizing that less could actually be more?
Time for some summer reading? Let me introduce you to one of my favorites: A novel about friendship, creativity, and the video games we build.
phdbytes.substack.com/p/tomorrow-a...
phdbytes.substack.com/p/tomorrow-a...
June 7, 2025 at 8:52 AM
Time for some summer reading? Let me introduce you to one of my favorites: A novel about friendship, creativity, and the video games we build.
phdbytes.substack.com/p/tomorrow-a...
phdbytes.substack.com/p/tomorrow-a...
In my latest post: Even with the most advanced AI assistant at your side, secure software still relies on human intervention.
AI Code Assistants: Fast, Helpful — and Dangerously Insecure?
TLDR; Even with the most advanced AI assistant at your side, secure software still relies on human intervention.
phdbytes.substack.com
June 4, 2025 at 6:04 AM
In my latest post: Even with the most advanced AI assistant at your side, secure software still relies on human intervention.
This week, I’m attending SWITS, the Swedish IT Security Research Network for Ph.D. students. I presented work on a study exploring the intersection of AI and cybersecurity.
June 3, 2025 at 1:21 PM
This week, I’m attending SWITS, the Swedish IT Security Research Network for Ph.D. students. I presented work on a study exploring the intersection of AI and cybersecurity.
Whether you work in cybersecurity, defense, policy, AI, or want to understand why headlines about ransomware and critical infrastructure attacks are becoming more common, this is a book well worth your time.
This Is How They Tell Me the World Ends
Nicole Perlroth (2021)
open.substack.com
June 1, 2025 at 7:19 AM
Whether you work in cybersecurity, defense, policy, AI, or want to understand why headlines about ransomware and critical infrastructure attacks are becoming more common, this is a book well worth your time.
📚🤖 Do LLMs abstract language or memorize copyrighted books?
In a substack post, I dive into a 101-page (!) paper showing how some models can reproduce entire novels.
An eye-opener for anyone working in AI, policy, or copyright.
In a substack post, I dive into a 101-page (!) paper showing how some models can reproduce entire novels.
An eye-opener for anyone working in AI, policy, or copyright.
When the Model _Is_ the Copy
A new study shows some LLMs don’t just learn from books — they remember them.
phdbytes.substack.com
May 29, 2025 at 7:11 AM
📚🤖 Do LLMs abstract language or memorize copyrighted books?
In a substack post, I dive into a 101-page (!) paper showing how some models can reproduce entire novels.
An eye-opener for anyone working in AI, policy, or copyright.
In a substack post, I dive into a 101-page (!) paper showing how some models can reproduce entire novels.
An eye-opener for anyone working in AI, policy, or copyright.
New editorial column in Corren this evening (Swedish):
I write about the real legacy of Linköping University: how it shaped lives, created friendships, and became a second home for thousands. Mine included.
👉 LiU’s 50-year jubilee is an ongoing promise of a future.
I write about the real legacy of Linköping University: how it shaped lives, created friendships, and became a second home for thousands. Mine included.
👉 LiU’s 50-year jubilee is an ongoing promise of a future.
CHRISTIAN GUSTAVSSON: På LiU träffar du den viktigaste i livet – personen du vill vara
Om Linköpings universitet har betytt mycket för vår region, så har det betytt ännu mer för oss människor.
www.corren.se
May 25, 2025 at 7:55 PM
New editorial column in Corren this evening (Swedish):
I write about the real legacy of Linköping University: how it shaped lives, created friendships, and became a second home for thousands. Mine included.
👉 LiU’s 50-year jubilee is an ongoing promise of a future.
I write about the real legacy of Linköping University: how it shaped lives, created friendships, and became a second home for thousands. Mine included.
👉 LiU’s 50-year jubilee is an ongoing promise of a future.
💡 Deep Work by Cal Newport might be the most important book for anyone doing serious thinking in a distracted world. I wrote about why it matters — and how to reclaim your focus.
Why Deep Work Might Be the Most Important Skill You’re Not Practicing
Cal Newport, 2016
phdbytes.substack.com
May 24, 2025 at 7:27 AM
💡 Deep Work by Cal Newport might be the most important book for anyone doing serious thinking in a distracted world. I wrote about why it matters — and how to reclaim your focus.
What comes after large language models? 🧠✨
Meta’s research into large concept models (LCMs) shifts focus from words to ideas, enabling reasoning across languages, formats, and long documents. I wrote a Substack post about why this concept might be a glimpse into the next frontier of AI.
Meta’s research into large concept models (LCMs) shifts focus from words to ideas, enabling reasoning across languages, formats, and long documents. I wrote a Substack post about why this concept might be a glimpse into the next frontier of AI.
Thinking in Concepts: A New Frontier for AI
Can Large Concept Models make machines think more like humans?
phdbytes.substack.com
May 20, 2025 at 8:53 AM
What comes after large language models? 🧠✨
Meta’s research into large concept models (LCMs) shifts focus from words to ideas, enabling reasoning across languages, formats, and long documents. I wrote a Substack post about why this concept might be a glimpse into the next frontier of AI.
Meta’s research into large concept models (LCMs) shifts focus from words to ideas, enabling reasoning across languages, formats, and long documents. I wrote a Substack post about why this concept might be a glimpse into the next frontier of AI.
In my latest Substack post, I explore a fascinating study on why we rarely act, even when we’re upset about data breaches. What holds us back, and how can we fix it?
phdbytes.substack.com/p/i-was-brea...
phdbytes.substack.com/p/i-was-brea...
May 14, 2025 at 10:32 AM
In my latest Substack post, I explore a fascinating study on why we rarely act, even when we’re upset about data breaches. What holds us back, and how can we fix it?
phdbytes.substack.com/p/i-was-brea...
phdbytes.substack.com/p/i-was-brea...
Editorial column out this morning in Corren (🇸🇪):
When Stångåstaden switches fiber provider, they don’t want tenants to know how money changes hands in Telia’s new deal. I write about a procurement that could cost us in Linköping – and violates key principles for how public companies should operate.
When Stångåstaden switches fiber provider, they don’t want tenants to know how money changes hands in Telia’s new deal. I write about a procurement that could cost us in Linköping – and violates key principles for how public companies should operate.
CHRISTIAN GUSTAVSSON: Har Stångåstaden glömt bort att de ägs av oss Linköpingsbor?
Har Stångåstaden tappat fotfästet och glömt bort att de ägs av oss Linköpingsbor?
www.corren.se
May 12, 2025 at 7:47 AM
Editorial column out this morning in Corren (🇸🇪):
When Stångåstaden switches fiber provider, they don’t want tenants to know how money changes hands in Telia’s new deal. I write about a procurement that could cost us in Linköping – and violates key principles for how public companies should operate.
When Stångåstaden switches fiber provider, they don’t want tenants to know how money changes hands in Telia’s new deal. I write about a procurement that could cost us in Linköping – and violates key principles for how public companies should operate.
Earlier this spring, I gave a short talk on using Generative AI for Cybersecurity. On Substack, I'm sharing some of the papers I read, current use-cases, and possible opportunities. phdbytes.substack.com/p/generative...
May 9, 2025 at 9:43 AM
Earlier this spring, I gave a short talk on using Generative AI for Cybersecurity. On Substack, I'm sharing some of the papers I read, current use-cases, and possible opportunities. phdbytes.substack.com/p/generative...
Five Steps Toward a Safer Digital Life – practical advice to strengthen your digital security.
Five Steps: Safer Digital Life
We can never be completely safe, but we can make it really difficult for others to claim our digital lives. Here are my five steps for becoming safer.
phdbytes.substack.com
May 6, 2025 at 5:27 PM
Five Steps Toward a Safer Digital Life – practical advice to strengthen your digital security.