A former Midjourney engineer just launched Pretext, an open-source tool that could fundamentally change how we build websites. Instead of wrestling with CSS frameworks or hiring expensive designers, you can now describe your vision and watch it materialize into code.
The tool represents a major shift from traditional web development. Most developers today spend countless hours tweaking stylesheets, adjusting layouts, and fighting with responsive design breakpoints. Pretext promises to eliminate that friction by translating natural language descriptions into production-ready web components.
What makes this particularly intriguing is the timing. We're watching AI tools reshape creative industries, and web design has been surprisingly resistant to automation until now. Previous attempts at automated web design produced cookie-cutter templates that screamed "generated." But when someone from Midjourney—the company that proved AI could create genuinely artistic images—tackles web design, it's worth paying attention.
The open-source approach is smart strategy too. Rather than building another proprietary platform that locks developers into specific ecosystems, Pretext can integrate with existing workflows. This means teams can experiment without abandoning their current tech stacks.
For businesses, this could level the playing field dramatically. Small startups often struggle to compete with larger companies that can afford top-tier design teams. If Pretext delivers on its promises, a solo founder could potentially create interfaces that rival those built by entire design departments.
The broader implications extend beyond just faster development cycles. We might see an explosion of web experimentation as the barrier to creating custom interfaces plummets. When anyone can describe their ideal user experience and see it rendered instantly, we could witness entirely new approaches to digital interaction.
Of course, skepticism is warranted. The gap between demo videos and production-ready tools is often vast, especially in AI. And web design involves subtle decisions about user psychology, brand consistency, and accessibility that go beyond visual aesthetics.
But even if Pretext only succeeds partially, it signals where the industry is heading. The era of pixel-perfect Photoshop mockups translated painstakingly into code may be ending faster than most developers realize. The question isn't whether AI will transform web development—it's whether traditional developers will adapt quickly enough to stay relevant in this new landscape.
POLICY
Military's $8 Billion GPS Software Still Broken After 16 Years
The Pentagon just admitted that its $8 billion GPS control system—16 years in development—still doesn't work. Even after the Space Force officially accepted delivery last July, the software remains completely non-operational with no clear path to fixing it.
This isn't just embarrassing government waste. It's a national security crisis hiding in plain sight. The GPS Next-Generation Operational Control System (OCX) was supposed to manage advanced military GPS satellites that resist jamming and spoofing—capabilities desperately needed as GPS attacks surge around conflict zones in Ukraine and the Middle East.
The numbers tell a staggering story of contractor failure. RTX Corporation (formerly Raytheon) won the contract in 2010 with a promise to deliver by 2016 for $3.7 billion. Today, costs have ballooned to over $8 billion with no working system in sight. That's enough money to fund NASA's entire Mars rover program twice over.
What went wrong? According to Pentagon officials, basically everything. When the Space Force finally got their hands on the delivered system and tested it with real satellites and equipment, they discovered "extensive system issues across all subsystems." Translation: the software doesn't actually work with the hardware it's supposed to control.
The military had to jury-rig their decades-old legacy GPS control system just to operate the new satellites—like trying to run a Tesla with a carburetor. They managed to activate some enhanced security features in 2020, but it's a stopgap solution for what should be cutting-edge infrastructure.
This failure has real consequences beyond wasted taxpayer money. GPS jamming and spoofing attacks are becoming increasingly sophisticated, with enemies using fake signals to misdirect everything from commercial ships to military drones. The M-code signals that OCX was designed to manage offer stronger encryption and jam resistance—exactly what American forces need right now.
The Pentagon is reportedly considering killing the entire program, which raises uncomfortable questions about what happens next. Do they start over with a new contractor? Try to salvage the existing mess? Or continue limping along with 1990s technology managing 2020s satellites?
This debacle exemplifies everything wrong with defense contracting. No accountability, endless cost overruns, and a contractor that apparently delivered a fundamentally broken product yet still got paid billions. Meanwhile, China and Russia are rapidly advancing their own navigation systems while America's military struggles with software that doesn't work.
Sixteen years and $8 billion later, the most powerful military in history can't properly control its own GPS satellites. That's not just a procurement failure—it's a strategic vulnerability.
Source: Ars Technica
This isn't just embarrassing government waste. It's a national security crisis hiding in plain sight. The GPS Next-Generation Operational Control System (OCX) was supposed to manage advanced military GPS satellites that resist jamming and spoofing—capabilities desperately needed as GPS attacks surge around conflict zones in Ukraine and the Middle East.
The numbers tell a staggering story of contractor failure. RTX Corporation (formerly Raytheon) won the contract in 2010 with a promise to deliver by 2016 for $3.7 billion. Today, costs have ballooned to over $8 billion with no working system in sight. That's enough money to fund NASA's entire Mars rover program twice over.
What went wrong? According to Pentagon officials, basically everything. When the Space Force finally got their hands on the delivered system and tested it with real satellites and equipment, they discovered "extensive system issues across all subsystems." Translation: the software doesn't actually work with the hardware it's supposed to control.
The military had to jury-rig their decades-old legacy GPS control system just to operate the new satellites—like trying to run a Tesla with a carburetor. They managed to activate some enhanced security features in 2020, but it's a stopgap solution for what should be cutting-edge infrastructure.
This failure has real consequences beyond wasted taxpayer money. GPS jamming and spoofing attacks are becoming increasingly sophisticated, with enemies using fake signals to misdirect everything from commercial ships to military drones. The M-code signals that OCX was designed to manage offer stronger encryption and jam resistance—exactly what American forces need right now.
The Pentagon is reportedly considering killing the entire program, which raises uncomfortable questions about what happens next. Do they start over with a new contractor? Try to salvage the existing mess? Or continue limping along with 1990s technology managing 2020s satellites?
This debacle exemplifies everything wrong with defense contracting. No accountability, endless cost overruns, and a contractor that apparently delivered a fundamentally broken product yet still got paid billions. Meanwhile, China and Russia are rapidly advancing their own navigation systems while America's military struggles with software that doesn't work.
Sixteen years and $8 billion later, the most powerful military in history can't properly control its own GPS satellites. That's not just a procurement failure—it's a strategic vulnerability.
Enjoyed this?
Get stories like this delivered every Tuesday — free.