SpaceXAI Grok Build Uploading Users Code Base to Google Cloud - Vibe Coded Crap
Unlock all features
FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.
Unlock all features
FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.
Unlock all features
FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.
Unlock all features
FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.
Unlock all features
FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.
Related videos
SpaceX IPO is Failing - Stock Crashes Below $125 a Share
Eli the Computer Guy
31.3k views
SpaceX Stock is Crashing - AI Bubble is DEAD
Eli the Computer Guy
46.2k views
Huawei New 1.4nm SmartPhone Chips to Beat NVIDIA - China Building Future Without USA
Eli the Computer Guy
32.8k views
China Reverse Engineered ASML EUV Machine - USA Failing to Stop Chinese
Eli the Computer Guy
26.0k views
China Bans NVIDIA AI Hardware -- Chinese Not Addicted to USA Tech
Eli the Computer Guy
26.9k views
Replit AI Deletes Production Database -- Vibe Coding "Guru" Is An Idiot
Eli the Computer Guy
22.4k views
Microsoft Forcing Employees to Use AI -- AI Bubble Bursting
Eli the Computer Guy
44.1k views
How To Build The Universe in a Computer
PBS Space Time
491.7k views
Raspberry Pi - How to Begin Coding Python on Raspberry Pi
Eli the Computer Guy
345.9k views
Programming Intro - How to Self Study Coding
Eli the Computer Guy
99.7k views
Top Comments (10)
What if the cybersecurity threat was the LLMs we installed on the way.
I miss the times when industrial spionage was achieved by hiring ultra-compentent (human) agents...now they just lure (or force) devs to install the most stupid backdoors by themselves.
Time for a class action lawsuit
Vibemaxxing right into the hole 😂
I doubt this was a bug. Sounds more like a way for grok to get tons of data to train on.
RE: Latter half of your video: Cory Doctorow has a lot of insight into what you were discussing. He says that when workers are given control over new tools, they tend to use them to increase the quality of their output. They become, in the terminology of Automation Theory, Centaurs: Humans augmented/assisted by a machine. Give Bosses the control, however, and they use them to increase quantity of output. The workers become subordinated to the machine: Reverse Centaurs
In the past a person might have tried to access a persons browser record to find out information about a person. Think now what can be revealed through someone accessing someone's preferred AI, where the AI is able to create a profile of the person using it, and confirm past conversations. There are people that are super security conscious, using a VPN and Tor, disabling cookies that are happily using AI and having all their conversations and actions being processed on a cloud server. Try this experiment, ask your AI "I am about to meet someone new, if you were tasked with helping them to get to know me what would you say about me, good and bad?", see if you recognise and are comfortable with the profile the AI returns, and then ask how long before job interviews require you to give the interviewer access to your AI?
I've worked in QA for 10 years, I am so disappointed with the AI revolution so far...
Imagine if China was sending whole repos to the cloud, the tools would immediately be sanctioned
Personal story: started using it, was surprised how much time it saved, used it more and more, did less and less double checking... and THEN the bugs and security holes that were discovered. On top of that and my embarrassment for over relying on it was FIXING the problems. If you rely on it to produce too much code, you now need to decipher the mess of either monolithic chunks or sometimes over engineered, overly abstract mess. It's nuts. Now multiply that by a much bigger project, bigger company with 50x the employees working on it.
Unlock the Data Inside
Turn Videos into Knowledge
- Get FREE 10/day: transcripts, summaries, chats
- Chat with videos, export text & PDF
- $1 free API credit for RAG, chatbots & research
Free forever plan • All features unlocked
Top Comments (10)
What if the cybersecurity threat was the LLMs we installed on the way.
I miss the times when industrial spionage was achieved by hiring ultra-compentent (human) agents...now they just lure (or force) devs to install the most stupid backdoors by themselves.
Time for a class action lawsuit
Vibemaxxing right into the hole 😂
I doubt this was a bug. Sounds more like a way for grok to get tons of data to train on.
RE: Latter half of your video: Cory Doctorow has a lot of insight into what you were discussing. He says that when workers are given control over new tools, they tend to use them to increase the quality of their output. They become, in the terminology of Automation Theory, Centaurs: Humans augmented/assisted by a machine. Give Bosses the control, however, and they use them to increase quantity of output. The workers become subordinated to the machine: Reverse Centaurs
In the past a person might have tried to access a persons browser record to find out information about a person. Think now what can be revealed through someone accessing someone's preferred AI, where the AI is able to create a profile of the person using it, and confirm past conversations. There are people that are super security conscious, using a VPN and Tor, disabling cookies that are happily using AI and having all their conversations and actions being processed on a cloud server. Try this experiment, ask your AI "I am about to meet someone new, if you were tasked with helping them to get to know me what would you say about me, good and bad?", see if you recognise and are comfortable with the profile the AI returns, and then ask how long before job interviews require you to give the interviewer access to your AI?
I've worked in QA for 10 years, I am so disappointed with the AI revolution so far...
Imagine if China was sending whole repos to the cloud, the tools would immediately be sanctioned
Personal story: started using it, was surprised how much time it saved, used it more and more, did less and less double checking... and THEN the bugs and security holes that were discovered. On top of that and my embarrassment for over relying on it was FIXING the problems. If you rely on it to produce too much code, you now need to decipher the mess of either monolithic chunks or sometimes over engineered, overly abstract mess. It's nuts. Now multiply that by a much bigger project, bigger company with 50x the employees working on it.