Michael Manueddu

Founder @wrldailabs @juiceOAS @exitsmokeai

[Michael Manueddu @Mike230313]

@opencode Yes thanks the service was stopping a lot yesterday, this morning a way better

[Jun Song @jun_song]

If Ox Alpha is GLM Flash,

It will be a frontier level AI running on DGX Spark.

[Michael Manueddu @Mike230313]

@opencode @grok model ranking?

[Michael Manueddu @Mike230313]

@MEMORYLOST731 @jun_song Yes exactly it's super suggestive and funny. Have fun building out there!

[Michael Manueddu @Mike230313]

@UnslothAI @Alibaba_Qwen May I cry? Or only devils do it?

[Pavel Durov @durov]

Telegram has applied for the .gram domain zone.

If the application is approved by ICANN, a billion Telegram users could get their own second-level domains — yourname.gram.

Users would be able to set up their interactive websites hosted by Telegram — with one prompt ✨

[Michael Manueddu @Mike230313]

@jun_song Im always complaining, telling on LinkedIn that Anthropic is a devil, they close us out only because they used us enough to train their super brains. I deleted Anthropic

[Jun Song @jun_song]

People freak out the second gas or egg prices go up even a tiny bit.

Yet AI keeps getting more expensive, and outside of myself, I barely see anyone complaining.

Even when API prices jump with every new release, or weekly usage caps keep shrinking. Cutting a limit in half is literally doubling the price.

AI is already becoming as essential to us as gas.

[Jacky Kwok @jackyk02]

Scaling self-verification with DeepSeek V4 Flash beats Claude Fable 5 on Terminal-Bench 2.1, while being 11x cheaper 💰

As open-source models become more capable, they can now generate large numbers of high-quality candidate solutions and verify their own outputs at very low cost.

For example, we find that sampling just 5 solutions with DeepSeek V4 Flash and ranking them using the same model with LLM-as-a-Verifier can lead to a significant boost in accuracy (79% → 88%), outperforming closed frontier models on Terminal-Bench.

Try it out today: https://t.co/UhudjKDaYI

More on verification scaling in my previous post.

[Jun Song @jun_song]

Stanford is definitely cooking something crazy.

By enabling self-verification during the intermediate process, Deepseek-V4-Flash easily outperforms Fable.

11x cheaper and a much higher Terminal bench score.

This is exactly why I love X. People just casually drop massive breakthroughs in a single post.

Github ⬇️

[Jun Song @jun_song]

If you truly understand how important AI is and the impact it’s going to have, you’d probably drop everything in your life and go all-in on it.

​Most people still don’t realize what’s coming.

[Jun Song @jun_song]

Seriously it's over.

This is not a good strategy imo.

No wonder why all executives left OAI recently.

Good luck with your IPO, Scam Altman.

[Michael Manueddu @Mike230313]

@MEMORYLOST731 @jun_song Thank you so much for helping, I'll try a model fit repo first I saw one yesterday, honestly I need to Develop more time and more things I have in mind, thanks I'll try 🙏🏼

[Michael Manueddu @Mike230313]

@0xLalice @jun_song He's totally right, the devil yes

[Michael Manueddu @Mike230313]

@zcode_ai Hello i login First time today some hour ago, but haven't received nothing @zcode_ai how I redeem them?

[Michael Manueddu @Mike230313]

@jun_song You are with the good guys good to know, nwo

[Jun Song @jun_song]

Paid KOLs for OpenAI and Anthropic probably have no clue how messed up what they're doing is.

You're literally siding with the devil and licking their boots to destory the entire humanity.
Consider dropping those contracts.

[Superman @thesupermanmx]

Two economists mathematically proved that AI will destroy the economy.

Researchers from Wharton and Boston University published a terryfiying paper called "The AI Layoff Trap."

They mapped out the economic end-game of the AI transition, and it exposes a fatal flaw in competitive capitalism.

When a company replaces a worker with AI, it captures 100% of the wage savings.

But that displaced worker is also a consumer. When they lose their job, they stop buying things.

The company gets all the savings, but the loss of consumer demand is spread across the entire economy.

If there are 20 competitors in a market, a CEO only absorbs 1/20th of the economic damage their layoffs just created.

So every single rational CEO has a mathematical incentive to automate as fast as possible.

They can literally see the cliff approaching, and they still step on the gas.

It triggers an unavoidable Prisoner’s Dilemma. If you don't automate, your competitors will, and they will crush you on price.

It doesn't just hurt workers. It destroys the businesses, too.

The economy gets trapped in an automation arms race. Companies fire their workforce to stay competitive, until the entire consumer base is completely hollowed out.

At the limit, the paper concludes: “Firms automate their way to boundless productivity and zero demand.”

And the scariest part?

The researchers mathematically tested every popular fix.

Universal Basic Income? Fails. It raises the living standard but doesn't change the corporate incentive to cut jobs. Retraining? Fails. Worker equity? Fails.

The paper proves that more competition actually makes the collapse happen faster. And "better" AI makes the damage worse.

The only thing that mathematically stops the collapse is a targeted automation tax, forcing companies to pay for the purchasing power they destroy before they automate the job.

[Jun Song @jun_song]

Local AI hardware guide :

~$2k : rtx3090 desktop with Qwen3.8-27b
~$4k : DGX Spark with Deepseek-V4-Flash 2.5bit
~$7k : Macbook Pro m5 max 128gb with Deepseek-V4-Flash 2.5bit
~$8k : 2xDGX Spark with nvfp4 or 6.5bit DQ
~$12k : 3xDGX Spark with GLM-5.3 3bit DQ