frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

AI future is more egalitarian

2•shoman3003•1h ago
I was reading Machiavelli where he argues that "freedom in a republic literally needs class conflict". Which is of course completely true no matter how much our current culture hates it to be.

There is a huge difference between founders and employees for example, founders don't really care about comfort whereas suits life goal is comfort (which is why you shouldn't even fund a founder who cares about restaurant/car/hotel like you do). Same in society, the elites want power/legacy above all, but the middle class just wants to add a little more comfort to their life. You can see that in the eyes of anyone you talk to, what their end goal is just as easily as you can see it in the eyes if a particular animal is a wolf or a sheep.

If the sheep won the culture war, our society will certainly be better short term. There will be more resources spent on welfare, less on the military. You can bet your ass there will be a UBI circling around to the point some of the weaker founders might see no value anymore in building a startup when you already have a good future guaranteed for all. The problem with this type of future scenario is that it's almost guaranteed the region where this happens will end up poorer and less technological. This happened to my civilization in the last century after centuries of golden ages, and it will happen to yours just as easily if you let comfort win.

Okay, now that we established that societies do actually need class struggle to maintain freedom. Would the same thing happen to a virtual society, whether it was made of AIs or mapped brains? The answer is simple, %100 YES. Why? because compute is not easily available.

Let's imagine a future 100 years from now where most humans are using a Neuralink to live in a virtual social media shthole where AIs are made to be the servant class. The humans are not all equal in that world, because the more compute you can acquire the more power you have & the more you can innovate. Some people will own shares in the system (trillionaires), others will have lots of compute at their disposal (elites), while others will only have the basic amount of compute necessary to survive and have some fun (they are kinda comfortable playing simulations all day).

Let's compare that reality to our own:

Bits: vHumans have almost unlimited health insurance, luxurious houses, the unlimited food/drinks. Atoms: The wealthiest nations are barely able to keep people out of the streets, even Norway is not wealthy enough to distribute $10k/m to everyone.

Bits: VHs can easily acquire more compute if they had the will to escape playing around in simulations all day (which, let's be honest, is not easy. I can barely stop myself from playing civ) Atoms: Humans need VC money and years of living on noodles (putting it nicely) to have %10 chance at most to move from the middle class up. Many billionaires even had to spend a couple of years on the streets, if that doesn't prove it, I don't know what will.

Bits: Entertainment for the middle class is incredible. You can actually move back and forward in time and experience it all. Atoms: Entertainment is a punch of sh

tty tiktoks that are editable but deeply unsatisfying.

Atoms: Private equity is very hard to measure, track and keep accountable. Bits: You can see how much compute everyone has on your dashboard, playing in the shadows is much harder.

Simple, standalone tools for working with multimodal data

https://github.com/mixpeek/multimodal-tools
1•Beefin•41s ago•0 comments

AWS on X402 for Agentic Commerce

https://aws.amazon.com/blogs/industries/x402-and-agentic-commerce-redefining-autonomous-payments-...
1•bethecloud•1m ago•0 comments

Docker Containers: An Interactive Explanation

https://devopsagents.co/blog/containers-from-scratch
3•jcyriac•1m ago•0 comments

Trump Airs a House Republican's Terminal Diagnosis, Claiming to Have Reversed It

https://www.nytimes.com/2026/03/16/us/politics/trump-neal-dunn-terminal-diagnosis-johnson.html
1•duxup•5m ago•1 comments

A reminder to check your inactive account passwords

https://shub.club/writings/2026/march/rotate-your-passwords/
1•forthwall•5m ago•0 comments

The Last Quiet Thing

https://www.terrygodier.com/the-last-quiet-thing
2•JumpCrisscross•6m ago•0 comments

As India seeks Hormuz safe passage, Tehran asks for return of seized tankers

https://www.reuters.com/world/india/india-seeks-hormuz-safe-passage-tehran-asks-return-seized-tan...
1•JumpCrisscross•7m ago•0 comments

Robot dogs are protecting data centers. Operators are seeing payoffs

https://www.businessinsider.com/robot-dogs-quadruped-data-center-security-boston-dynamics-ghost-r...
1•mikhael•7m ago•0 comments

Show HN: Aelitium – Git-style verification for LLM outputs

https://aelitium.com
1•catarina_eng•8m ago•0 comments

Ask HN: Okara's new AI CMO: Will this solve the marketing problem?

1•nazbasho•10m ago•1 comments

"Gemini AOS": Google's Next Android, Hidden in ATI Silicon

https://guanghuimao.substack.com/p/gemini-aos-googles-next-android-hidden
2•Ati985•11m ago•0 comments

Avalonia WebView Is Now Free and Open Source

https://github.com/AvaloniaUI/Avalonia.Controls.WebView
2•sltr•11m ago•0 comments

Google scraps AI search feature that crowdsourced amateur medical advice

https://www.theguardian.com/technology/2026/mar/16/google-scraps-ai-search-feature-that-crowdsour...
2•geox•11m ago•0 comments

Video Conferencing with Postgres

https://planetscale.com/blog/video-conferencing-with-postgres
1•thunderbong•11m ago•0 comments

tables and views

https://web.archive.org/web/20060710204054/http://www.kx.com/a/k/examples/table.k
1•tosh•12m ago•0 comments

Show HN: Ziex – A full-stack web framework for Zig

2•nurulhudaapon•13m ago•0 comments

How we Built Private Post-Training and Inference for Frontier Models

https://www.workshoplabs.ai/blog/private-post-training
4•oscarmoxon•15m ago•1 comments

Show HN: SmartVPN: A VPN That Hides in Plain Sight over WebSockets

https://typescript.guru/smartvpn-a-vpn-that-hides-in-plain-sight-over-websockets/
1•PhilKunz•16m ago•0 comments

Show HN: LLM Memory Storage that scales, easily integrates, and is smart

https://github.com/colinulin/mind-palace
2•pocketcolin•18m ago•0 comments

Agents over Bubbles

https://stratechery.com/2026/agents-over-bubbles/
2•jonbaer•19m ago•0 comments

Lines of Code as a Productivity Metric in the AI Era

https://keegan.codes/blog/lines-of-code-as-a-productivity-metric-ai-era
1•keegandonley•20m ago•0 comments

Show HN: Stop renting AI – run AI workers on your own dedicated node

https://ainode.sh
1•rzessski•22m ago•1 comments

Northstar CUA Fast, open source 4B CUA model

https://www.tzafon.ai/blog/northstar-cua-fast
1•publius_•23m ago•0 comments

How Tenaciously Palantir Courted Switzerland

https://www.republik.ch/2026/02/18/how-tenaciously-palantir-courted-switzerland
2•sschueller•25m ago•0 comments

Beside

https://aplwiki.com/wiki/Beside
1•tosh•25m ago•0 comments

2 Days to Ship: Codex-Authored Metal Compute Shaders in Draw Things

https://engineering.drawthings.ai/p/2-days-to-ship-codex-authored-metal
1•liuliu•26m ago•0 comments

Open Decision-Making (2021)

https://web.stanford.edu/~ouster/cgi-bin/decisions.php
1•otoolep•26m ago•0 comments

Show HN: Tic-Tac-Word – Can you beat yourself in this tic-tac-toe word game?

https://www.tictacword.com
4•onion92•27m ago•1 comments

Autoresearch for SAT Solvers

https://github.com/iliazintchenko/agent-sat
2•chaisan•28m ago•1 comments

Agents are not thinking: a behavioral study of pwning sonnet

https://technoyoda.github.io/pwning-claude.html
2•sci-genie•28m ago•0 comments