Hacker News Reader

DebtDeflationon Generative AI Hype Peaking

I stop reading/listening as soon as AGI or Superintelligence is mentioned.

>forces participants to consider long term outcomes more

I remember many years ago telling a senior executive that I had concerns that some of the steps we were taking to boost current quarter financials would negatively impact business performance in the long term. He just chuckled and said, "there's no such thing as the long term, only a never ending series of current quarters".

DebtDeflationon Layoffs Don't Work

They used to be a once a decade "save the company during a recession" move. Now they seem to be a quarterly "manage earnings per share" move.

DebtDeflationon Microsoft is plotting a future without OpenAI

https://techcrunch.com/2025/03/05/openai-reportedly-plans-to...

It points to an article on "The Information" as the source, but that link is paywalled.

DebtDeflationon Microsoft is plotting a future without OpenAI

Again, irrelevant. We're talking about orders of magnitude here. Current pricing is in line with most SaaS pricing - tens of dollars to hundreds of dollars per seat per month. Now they're suddenly talking about thousands of dollars to tens of thousands of dollars per seat per month.

DebtDeflationon Microsoft is plotting a future without OpenAI

The "3-5" is certainly overstated, but you definitely can hire ONE PhD for that price, just as you can hire a SWE for $120K or a knowledge worker for $24K. The point is that from a CEO's perspective "replacing all the humans with AI" looks a lot less compelling when the AI costs the same as a human worker or even a significant fraction of a human worker.

DebtDeflationon Microsoft is plotting a future without OpenAI

A couple of days ago it leaked that OpenAI was planning on launching new pricing for their AI Agents. $20K/mo for their PhD Level Agent, $10K/mo for their Software Developer Agent, and $2K/mo for their Knowledge Worker Agent. I found it very telling. Not because I think anyone is going to pay this, but rather because this is the type of pricing they need to actually make money. At $20 or even $200 per month, they'll never even come close to breaking even.

DebtDeflationon Simple Explanation of LLMs

Would love to see a similar explanation of how "reasoning" versions of LLMs are trained. I understand that OpenAI was mum about how they specifically trained o1/o3 and that people are having to reverse engineer from the DeepSeek paper which may or may not be a different approach, but would like to see a coherent explanation which is not just an regurgitation of Chain of Thought or handwavy "special reasoning tokens give the model more time to think".

DebtDeflationon Scientist Left OpenAI Last Year. His Startup Is Already Worth $30B

SSI has no product and has publicly stated that they will not release any products until they have developed AGI which will be their first and only product. This is somehow worth $30B.

DebtDeflationon Microsoft's new Dragon Copilot is an AI assistant for healthcare

This sounds like a basic STT/Transcription app. What makes it a "Healthcare Virtual Assistant"? Presumably it's been trained on a medical dictionary to recognize vocabulary from this domain? Dragon has been making transcription apps since 1997, originally based on Hidden Markov Models, I assume since updated to use transformers.

DebtDeflationon Schools reviving shop class

Was going to say. I graduated high school 33 years ago. Was one of the last cohorts that had wood shop and metal shop. It was fun, but even back then the tools and presses we used hadn't been relevant to industry for awhile. If you want shop to be relevant to modern industry you would need to teach robotics, CAD/CAM, CNC, 3D printing, etc.

DebtDeflationon GPT-4.5

If you read what sama is quoted as saying in your link, it's obvious that "unified model" = router.

> “We hate the model picker as much as you do and want to return to magic unified intelligence,”

> “a top goal for us is to unify o-series models and GPT-series models by creating systems that can use all our tools, know when to think for a long time or not, and generally be useful for a very wide range of tasks,”

> the company plans to “release GPT-5 as a system that integrates a lot of our technology, including o3,”

He even slips up and says "integrates" in the last quote.

When he talks about "unifying", he's talking about the user experience not the underlying model itself.

DebtDeflationon GPT-4.5

GPT 5 is likely just going to be a router model that decides whether to send the prompt to 4o, 4o mini, 4.5, o3, or o3 mini.

DebtDeflationon Larry Ellison's half-billion-dollar quest to change farming

At least someone got the reference, thank you.

DebtDeflationon Slack Is Down

I can't access the desktop or web versions but my phone app still seems to be working, at least intermittently.

DebtDeflationon Larry Ellison's half-billion-dollar quest to change farming

As long as you have a dozen potatoes, some human poop, a sample of earth soil for necessary nutrients and bacteria, and rocket fuel to burn to make water, it should be pretty easy.

DebtDeflationon CDC terminates flu vaccine promotion campaign

As recently as a year ago, the right wing vaccine opposition was focused almost exclusively on mRNA Covid vaccines. This sudden opposition to ALL vaccines, from Polio to Flu vaccines, dates back to when RFK announced he was suspending his campaign and joining Trump or shortly thereafter.

DebtDeflationon Broken Legs and Ankles Heal Better If You Walk on Them Within Weeks

I have some books translated from Russian that that I acquired back in my powerlifting days and that showed studies the Russians did on Olympic weightlifters. The children (who had been lifting heavy since a very young age) had bone mineral densities far higher than average untrained adults. The adult Olympic weight lifters had bone mineral densities that were completely off the scale, as the book put it "in polar bear territory". Granted, these study participants were people explosively lifting hundreds of pounds from floor to overhead in the blink of an eye, not your average gymbro, but still.

DebtDeflationon Why Quantum Cryptanalysis is Bollocks [pdf]

Experiments to show that solid-state materials were capable of amplifying or switching one electric voltage/current with another.

I'm not disputing that these later became the building blocks for computers, just as the quantum gate experiments being conducted in laboratories today may become the building blocks for quantum computers in the future. But that's not where we are now.

DebtDeflationon Why Quantum Cryptanalysis is Bollocks [pdf]

Were the first point-contact transistors that were developed in the 1940s at Bell Labs and that consisted of a piece of gold foil stretched over a plastic wedge which was then cut by hand at the tip and pressed against a block of germanium stacked on a metal base, a computer or an experiment?

DebtDeflationon The Generative AI Con

>People like Marc Benioff claiming that "today's CEOs are the last to manage all-human workforces"

This is the real problem. Companies have ALREADY starting laying off significant percentages of their workforce because they're buying into the AI "digital worker" hype without any idea of how exactly AI is going to do the jobs of 80% of THEIR employees across all departments in the next year or so.

DebtDeflationon Why Quantum Cryptanalysis is Bollocks [pdf]

>"These are physics experiments, not computers. Claiming that it’s a computer misrepresents what we’re really working with. Every time you see “quantum computer” mentally substitute “physics experiment”, which is what’s actually being discussed."

This needs to be a disclaimer on every discussion of "quantum computing."

DebtDeflationon Plane crashes, overturns during landing at Toronto airport

After touchdown, but it rolled sideways not end over end. This kept the fuselage intact and ripped off the wings (where all the fuel is) which is why everyone survived.

DebtDeflationon Perplexity Deep Research

There's a lot of mental inertia combined with an extremely fast moving market. Google was behind in the AI race in 2023 and a good chunk of 2024. But they largely caught up with Gemini 1.5, especially the 002 release version. Now with Gemini 2 they are every bit as much of a frontier model player as OpenAI and Anthropic, and even ahead of them in a few areas. 2025 will be an interesting year for AI.

DebtDeflationon LM2: Large Memory Models

Just as Graph RAG should have just been GAG.

DebtDeflationon What happens to your online accounts when you die?

A guy I worked with at a former employer died around 2015. Every year, for almost a decade, I get a notification from LinkedIn to congratulate him on his work anniversary.

DebtDeflationon Nvidia's RTX 5090 power connectors are melting

Search Amazon for "240W USB" and you get multiple pages of results for cables.

A few years ago there was a recall of OnePlus cables that were melting and catching fire, I had 2 of them and both melted.

But yes 240W/48V/5A is insane for a spec that was originally designed for 0.5W/5V/100mA. I suspect this is the limit for USB charging as anything over 48V is considered a shock hazard by UL and 5A is already at the very top of the 3-5A limit of 20AWG for fire safety.

DebtDeflationon DeepScaleR: Surpassing O1-Preview with a 1.5B Model by Scaling RL

>Every other LLM I've tried include o3-mini-high: Fill the 12-liter jug completely. Pour it into the 6 liter jug.

Try it with a 12L jug and a 4L jug and ask for 4L. See if it tells you to just fill the 4L or to fill the 12L and pour into the 4L twice discarding both times and there will be 4L remaining in the 12L jug.

Even though it's still technically correct, it demonstrates that there's no real "reasoning" happening just regurgitation of training data.

DebtDeflationon TSMC 2nm Process Disclosure – How Does It Measure Up?

In late 2013 Intel said 10nm was coming in 2015.

DebtDeflationon Working from home is here to stay

And such a transparent one at that. Companies closed offices and significantly reduced footprints in offices that stayed open. Now they're ordering employees to return to offices that do not exist anymore or that do not have enough workspaces for them.