Mon · 13 Jul 2026·Issue 032
Decoded.
·Subscribe →
AI / INDEX
CLAUDE-CODESDK BILLING SPLITJUN 15EU-AI-ACTDEC 2027DELAYGUARD-ACTCOMMITTEEUNAN.FRONTIEREXPORT CTRLLIFTEDAI-SPEND$2.59T2026
UPDATED 06 JUL ET
Back to this week
Week ofApril 27 / 202605 stories / 3 bonus / ~27 min total

The Reading List.

Contents01OpenAI Releases GPT-5.5, Bringing Company...02AI Will Reshape More Jobs...03MIT Study Challenges AI Job...04China's AI Upstart DeepSeek Drops...05Tech Industry Lays Off Nearly...
01
/ LEADRead Sunday
Bucketmodels
LevelAccessible
SourceTechCrunch
Read5 min

OpenAI Releases GPT-5.5, Bringing Company One Step Closer to an AI Super App

GPT-5.5 launched April 23 with a clear shift toward agentic work: it can take a goal, break it into steps, and complete the whole task without step-by-step guidance. The gains are concentrated in coding, computer use, and knowledge work. The useful frame here is not the benchmark scores but the direction of travel: OpenAI is building toward a model that operates more like a junior employee than a search engine. That changes what "using AI at work" will mean within the next year or two.

Read on TechCrunch ->
# openai# gpt# agentic# models
02
Read Wednesday
Bucketbusiness
LevelAccessible
SourceBCG
Read8 min

AI Will Reshape More Jobs Than It Replaces

BCG projects that 10 to 15 percent of U.S. jobs could be eliminated within five years, while 50 to 55 percent will be reshaped rather than eliminated. Workers keep their jobs but face significantly different expectations for how they work. The report is worth reading alongside this week's layoff headlines because it provides the longer arc: displacement is real and concentrated in specific task types, but the dominant story for most professionals is role evolution, not role elimination.

Read on BCG ->
# workforce# jobs# future-of-work# research
03
Read Wednesday
Bucketbusiness
LevelAccessible
SourceAxios
Read4 min

MIT Study Challenges AI Job Apocalypse Narrative

MIT research pushes back on the sudden-cliff version of AI displacement: instead of mass replacement hitting all at once, the evidence points to a slower, uneven reshaping of work that varies significantly by role, industry, and employer. The practical takeaway is that "AI is taking jobs" and "AI is not taking jobs" are both too blunt to be useful. What matters is which specific tasks within a role are being automated, and on what timeline.

Read on Axios ->
# jobs# research# displacement# mit
04
Read Friday
Bucketmodels
LevelAccessible
SourceCNN Business
Read5 min

China's AI Upstart DeepSeek Drops New Model. Will It Make Waves Like Last Year?

DeepSeek released V4 on April 24, one year after its R1 model shook assumptions about who can build frontier AI. V4 supports a one-million token context window and runs on Chinese-made chips rather than Nvidia hardware, which matters because it demonstrates the AI race is no longer constrained by U.S. export controls. For non-technical readers, the key point is competition: more capable models from more players means the tools available to you improve faster, and costs come down.

Read on CNN Business ->
# deepseek# china# models# competition
05
Read Friday
Bucketbusiness
LevelAccessible
SourceTom's Hardware
Read5 min

Tech Industry Lays Off Nearly 80,000 Employees in Q1 2026 — Almost 50% Due to AI

78,557 tech workers lost jobs in Q1 2026, with roughly half of those cuts attributed to AI automation. Customer support has been the hardest-hit category. The article is worth reading alongside this week's other picks because the numbers are real but the framing is contested: a December 2025 survey found 59 percent of hiring managers admit they cite AI in layoff announcements because it "plays better with stakeholders" than acknowledging budget pressure. Both things are true at once, and separating them matters for how you read your own industry's headlines.

Read on Tom's Hardware ->
# layoffs# workforce# automation# tech

Bonus material

For the curious.

Optional / 03 reads
MIT Technology Review

Three Reasons Why DeepSeek's New Model Matters

MIT Tech Review goes beyond the benchmark comparisons to explain what V4 means for the global AI supply chain and chip independence. More analytical than the CNN piece.

Next Monday

Get the next issue on Monday.

One email. Five stories. Three bonus reads. A new career impact analysis. All free.