César Solórzano
@cesarsolorzano.bsky.social
Coding for living 👨💻
https://blog.bonfy.ai/the-end-of-meritocracy-ai-human-value-and-what-comes-next
t.co
October 13, 2025 at 2:40 PM
Reposted by César Solórzano
I think the best way to learn something is to build it from scratch. That’s why I love articles like “Building framework X from scratch.”
And here is a collection of many such articles compiled into one list on various topics.
github.com/codecrafters...
And here is a collection of many such articles compiled into one list on various topics.
github.com/codecrafters...
June 2, 2025 at 1:11 PM
I think the best way to learn something is to build it from scratch. That’s why I love articles like “Building framework X from scratch.”
And here is a collection of many such articles compiled into one list on various topics.
github.com/codecrafters...
And here is a collection of many such articles compiled into one list on various topics.
github.com/codecrafters...
Reposted by César Solórzano
Channel 7 News in Australia posted this clip as if it was perfectly normal
April 22, 2025 at 5:09 AM
Channel 7 News in Australia posted this clip as if it was perfectly normal
Reposted by César Solórzano
✨ In March 2023, I published an article about AI.
This was right after OpenAI showcased how GPT4 could turn a hand-drawn sketch of a website into real HTML/CSS/JS. The consensus online was that FE jobs would stop existing within 1-2 years.
Well, it’s been 2 years! So, I just published a follow-up:
This was right after OpenAI showcased how GPT4 could turn a hand-drawn sketch of a website into real HTML/CSS/JS. The consensus online was that FE jobs would stop existing within 1-2 years.
Well, it’s been 2 years! So, I just published a follow-up:
The Post-Developer Era • Josh W. Comeau
When OpenAI released GPT-4 back in March 2023, they kickstarted the AI revolution. The consensus online was that front-end development jobs would be totally eliminated within a year or two.Well, it’s ...
www.joshwcomeau.com
April 14, 2025 at 3:11 PM
✨ In March 2023, I published an article about AI.
This was right after OpenAI showcased how GPT4 could turn a hand-drawn sketch of a website into real HTML/CSS/JS. The consensus online was that FE jobs would stop existing within 1-2 years.
Well, it’s been 2 years! So, I just published a follow-up:
This was right after OpenAI showcased how GPT4 could turn a hand-drawn sketch of a website into real HTML/CSS/JS. The consensus online was that FE jobs would stop existing within 1-2 years.
Well, it’s been 2 years! So, I just published a follow-up:
Reposted by César Solórzano
ChatGPT Deep Research does indeed hallucinate when given a false premise (or when asked to do research on the ungrounded facts ChatGPT hallucinates). This is what I call "induced hallucination":
April 14, 2025 at 4:39 AM
ChatGPT Deep Research does indeed hallucinate when given a false premise (or when asked to do research on the ungrounded facts ChatGPT hallucinates). This is what I call "induced hallucination":
Reposted by César Solórzano
1. LLM-generated code tries to run code from online software packages. Which is normal but
2. The packages don’t exist. Which would normally cause an error but
3. Nefarious people have made malware under the package names that LLMs make up most often. So
4. Now the LLM code points to malware.
2. The packages don’t exist. Which would normally cause an error but
3. Nefarious people have made malware under the package names that LLMs make up most often. So
4. Now the LLM code points to malware.
LLMs hallucinating nonexistent software packages with plausible names leads to a new malware vulnerability: "slopsquatting."
LLMs can't stop making up software dependencies and sabotaging everything
: Hallucinated package names fuel 'slopsquatting'
www.theregister.com
April 12, 2025 at 11:43 PM
1. LLM-generated code tries to run code from online software packages. Which is normal but
2. The packages don’t exist. Which would normally cause an error but
3. Nefarious people have made malware under the package names that LLMs make up most often. So
4. Now the LLM code points to malware.
2. The packages don’t exist. Which would normally cause an error but
3. Nefarious people have made malware under the package names that LLMs make up most often. So
4. Now the LLM code points to malware.
Reposted by César Solórzano
EXCLUSIVE: Microsoft and OpenAI are investigating whether a group linked to China's DeepSeek obtained OpenAI's data.
Microsoft Probing If DeepSeek-Linked Group Improperly Obtained OpenAI Data
Microsoft Corp. and OpenAI are investigating whether data output from OpenAI’s technology was obtained in an unauthorized manner by a group linked to Chinese artificial intelligence startup DeepSeek, ...
www.bloomberg.com
January 29, 2025 at 3:17 AM
EXCLUSIVE: Microsoft and OpenAI are investigating whether a group linked to China's DeepSeek obtained OpenAI's data.
Reposted by César Solórzano
OpenAI complaining China’s DeepSeek steal their work that they stole from Google that they stole from artists, writers, and everyone just goes to show you how this has nothing to do with “innovation” and everything to do with domination.
January 27, 2025 at 9:42 PM
OpenAI complaining China’s DeepSeek steal their work that they stole from Google that they stole from artists, writers, and everyone just goes to show you how this has nothing to do with “innovation” and everything to do with domination.
Reposted by César Solórzano
Laughing my ass off at OpenAI drawing a line in the sand over the idea companies like DeepSeek might have used their data without consent.
"Don't steal what we stole" isn't a compelling argument.
"Don't steal what we stole" isn't a compelling argument.
January 29, 2025 at 9:39 AM
Laughing my ass off at OpenAI drawing a line in the sand over the idea companies like DeepSeek might have used their data without consent.
"Don't steal what we stole" isn't a compelling argument.
"Don't steal what we stole" isn't a compelling argument.
Reposted by César Solórzano
I think modern computer software was a mistake and we should go back to our roots of developing for limited hardware.
The same amount of electricity that could power MS-DOS for a year is now used to ask Copilot how to save a Word document to your local Desktop instead of OneDrive.
The same amount of electricity that could power MS-DOS for a year is now used to ask Copilot how to save a Word document to your local Desktop instead of OneDrive.
January 28, 2025 at 8:34 PM
I think modern computer software was a mistake and we should go back to our roots of developing for limited hardware.
The same amount of electricity that could power MS-DOS for a year is now used to ask Copilot how to save a Word document to your local Desktop instead of OneDrive.
The same amount of electricity that could power MS-DOS for a year is now used to ask Copilot how to save a Word document to your local Desktop instead of OneDrive.
Reposted by César Solórzano
"Does your dog bite?"
"No, it's worse... she judges"
"No, it's worse... she judges"
January 29, 2025 at 12:58 AM
"Does your dog bite?"
"No, it's worse... she judges"
"No, it's worse... she judges"
Reposted by César Solórzano
Beautiful morning light illuminates the Canadian Rockies' towering Mount Assiniboine as, ever so briefly, underlying rocks and pebbles show through before the winds pick up again.
#bluesky #photography #landscape #travel #nature #scape
#landscapephotography #naturephotography #sunrise #canada
#bluesky #photography #landscape #travel #nature #scape
#landscapephotography #naturephotography #sunrise #canada
January 27, 2025 at 12:22 PM
Beautiful morning light illuminates the Canadian Rockies' towering Mount Assiniboine as, ever so briefly, underlying rocks and pebbles show through before the winds pick up again.
#bluesky #photography #landscape #travel #nature #scape
#landscapephotography #naturephotography #sunrise #canada
#bluesky #photography #landscape #travel #nature #scape
#landscapephotography #naturephotography #sunrise #canada
Reposted by César Solórzano
A small Chinese artificial intelligence lab, DeepSeek, stunned the world by revealing the technical recipe for its cutting-edge model, turning its reclusive leader into a national hero.
Find out how the start-up shocked Silicon Valley: on.ft.com/4jvooqi
Find out how the start-up shocked Silicon Valley: on.ft.com/4jvooqi
January 27, 2025 at 12:43 PM
A small Chinese artificial intelligence lab, DeepSeek, stunned the world by revealing the technical recipe for its cutting-edge model, turning its reclusive leader into a national hero.
Find out how the start-up shocked Silicon Valley: on.ft.com/4jvooqi
Find out how the start-up shocked Silicon Valley: on.ft.com/4jvooqi
Reposted by César Solórzano
Breaking News: President Trump announced a barrage of tariffs and sanctions targeting Colombia after it refused to accept military deportation flights.
Trump Hits Colombia With Tariffs in Feud Over Military Deportation Flights
Gustavo Petro said on X that the United States should not treat Colombian migrants as criminals and that he had already turned away U.S. military flights carrying deportees.
www.nytimes.com
January 26, 2025 at 8:02 PM
Breaking News: President Trump announced a barrage of tariffs and sanctions targeting Colombia after it refused to accept military deportation flights.
After reading about Bluesky and Mastodon, I’ve got a few thoughts on how decentralization and data ownership in social networks serve two distinct audiences.
January 26, 2025 at 6:26 PM
After reading about Bluesky and Mastodon, I’ve got a few thoughts on how decentralization and data ownership in social networks serve two distinct audiences.
Reposted by César Solórzano
imagine misaligning an icon and now its on 100,000 packets
January 26, 2025 at 3:23 PM
imagine misaligning an icon and now its on 100,000 packets
Reposted by César Solórzano
Signs that Audible operates like a cushy monopoly:
- Has the highest take rate (75% of what customers pay for an audiobook is Audible's to keep)
- Has 10x longer approval times for new audibooks
And so The Software Engineer's Guidebook audiobook remains available everywhere - except on Audible!
- Has the highest take rate (75% of what customers pay for an audiobook is Audible's to keep)
- Has 10x longer approval times for new audibooks
And so The Software Engineer's Guidebook audiobook remains available everywhere - except on Audible!
January 26, 2025 at 1:17 PM
Signs that Audible operates like a cushy monopoly:
- Has the highest take rate (75% of what customers pay for an audiobook is Audible's to keep)
- Has 10x longer approval times for new audibooks
And so The Software Engineer's Guidebook audiobook remains available everywhere - except on Audible!
- Has the highest take rate (75% of what customers pay for an audiobook is Audible's to keep)
- Has 10x longer approval times for new audibooks
And so The Software Engineer's Guidebook audiobook remains available everywhere - except on Audible!
Definitely a fun Saturday reading about AT Protocol and ActivityPub. I like the Fediverse, but the full network view offered by the AT Protocol has its advantages, particularly in user experience. UX can’t be overlooked if you want mainstream success. They have different objectives anyway.
January 26, 2025 at 5:38 AM
Definitely a fun Saturday reading about AT Protocol and ActivityPub. I like the Fediverse, but the full network view offered by the AT Protocol has its advantages, particularly in user experience. UX can’t be overlooked if you want mainstream success. They have different objectives anyway.
I’m definitely not mentioning Uruguay again, lol. It seems to be such a sensitive and polarized community.
January 26, 2025 at 4:35 AM
I’m definitely not mentioning Uruguay again, lol. It seems to be such a sensitive and polarized community.
From the start, letting Big Tech control speech was a huge mistake—it only pushed people to opposing extremes. The internet was meant to connect us, and I’m excited to see efforts to bring that vision back!
January 26, 2025 at 4:20 AM
From the start, letting Big Tech control speech was a huge mistake—it only pushed people to opposing extremes. The internet was meant to connect us, and I’m excited to see efforts to bring that vision back!
Por lo visto la comunidad en Uruguay de Bluesky está politizada, tomada por la izquierda radical y su resentimiento. Por qué polarizar una red social? Next!
January 26, 2025 at 3:45 AM
Por lo visto la comunidad en Uruguay de Bluesky está politizada, tomada por la izquierda radical y su resentimiento. Por qué polarizar una red social? Next!
Reposted by César Solórzano
I gave a talk today in which I walked through key points from our Bluesky architecture paper, and got some good questions www.cl.cam.ac.uk/research/sec...
Department of Computer Science and Technology – Security Group: 2024-11-22
www.cl.cam.ac.uk
November 22, 2024 at 8:20 PM
I gave a talk today in which I walked through key points from our Bluesky architecture paper, and got some good questions www.cl.cam.ac.uk/research/sec...
Today’s reading: “Bluesky and the AT Protocol: A Usable Decentralized Social Media Platform”
A must-read paper if you want to learn more about it!
arxiv.org/pdf/2402.03239
A must-read paper if you want to learn more about it!
arxiv.org/pdf/2402.03239
arxiv.org
January 26, 2025 at 1:28 AM
Today’s reading: “Bluesky and the AT Protocol: A Usable Decentralized Social Media Platform”
A must-read paper if you want to learn more about it!
arxiv.org/pdf/2402.03239
A must-read paper if you want to learn more about it!
arxiv.org/pdf/2402.03239