E
@eudtoxic
Are they still hiring only in certain US cities? It’s highly unlikely they’ll find the outstanding talent they want if they don’t hire globally
7
0
348
@GergelyOroszposted on X
They are paying this much for standout software engineers - who can generate better software, faster, using Claude Code vs “just” avg sw engineers (who do OK with it.) AI amplifies existing ability, and the best in any indusrty (who learn to use these tools) are in more demand
View original tweet on X →
Community Sentiment Analysis
Real-time analysis of public opinion and engagement
Sentiment Distribution
86% Engaged
72% Positive
Positive
72%
Negative
14%
Neutral
14%
Key Takeaways
What the community is saying — both sides
Supporting
1
AI raises the ceiling, not the floor:
top engineers get a much bigger productivity boost than mediocre ones, so the performance gap is widening.
2
AI amplifies excellence across roles:
sales and GTM are boosted too — AI multiplies top performers’ impact but doesn’t replace elite sellers or engineers.
3
Judgment is the scarce skill:
companies pay for knowing when AI is wrong — constraint design, failure-mode anticipation and system architecture are the premium capabilities.
4
Compensation and hiring will shift:
highly AI-productive engineers will command much higher pay, firms may restructure, and hiring globally becomes necessary to find top talent.
5
AI exposes coasting:
tools make it easier to identify who’s genuinely excellent and who has been coasting.
6
Bad fundamentals get scaled faster:
AI doesn’t fix sloppy architecture — it accelerates engineers who already know what to verify and can amplify poor practices if judgment is weak.
7
Small teams and solo founders gain leverage — but not the last mile:
tools like Claude Code can close ~80% of the gap to a small team, but the final ~20% — taste, judgment, and product decisions — remains human; the big question is what happens when that multiplier becomes ubiquitous.
Opposing
1
rare skill of finding where Claude’s reasoning breaks before shipping
and for running infra at a scale only a handful of firms touch.
2
day-to-day users rarely need model-breaking research or hyperscale infra
.
3
only amplify engineers who already know what the correct output should look like
; otherwise you just ship bugs at 10x speed.
4
brute-force code generation over engineering quality
prioritizing quick, working-looking outputs rather than long-term correctness and maintainability.
5
“software engineering is dead”
the claim doesn’t match how teams actually operate and is being questioned by practitioners.
Top Reactions
Most popular replies, ranked by engagement
D
@devruso
oh, so AI can’t replace good devs, huh?
6
0
258
R
@robinengg
I feel a big reset is coming. SWEs that are highly more productive with AI will be paid a lot more. Does Big Tech feel it's easier to lay off people for this readjustment, or not, is the big question.
3
1
611
C
@crepesupreme
Frontier lab eng pay isn't an amplification story. The rare skill they need is finding where Claude's reasoning breaks before it ships, and running infra at a scale maybe 5 firms in the world touch. Most teams using Claude Code never face either problem.
2
1
1.2K
V
@vinrambone
Who went on video and said software engineering was dead?
2
0
648
F
@ForbiddenSteve
counterpoint: Claude Code only amplifies you if you actually know what the output is supposed to look like, otherwise you just ship bugs at 10x speed 💀
1
1
500
Report an Issue
Found something wrong with this article? Let us know and we'll look into it.