HNHacker News
TopNewBestAskShowJobs

malshe

7,344 karma · joined December 27, 2017

submissionscomments
malshe··on Restrooms along the Norwegian Scenic Routes
Thanks! I was scrolling for this helpful comment with the link!
malshe··on US tech stocks see sharp falls over AI cost fears
> Intel reports PC sales are down. That's likely related to RAM costs, which made price/performance of new PCs drop.

My university has put all the laptop and PC purchases for staff and faculty on hold and they are approving them only on a case by case basis.

malshe··on MacBook Pro with PROLED XDR Display and Touch
Some of my students use touchscreen Windows laptops. They seem to be OK with them but I never found them usable particularly for coding. A touchscreen laptop has ergonomically bad design. Or it could just be a generational thing.
malshe··on MacBook Pro with PROLED XDR Display and Touch
Not an M6 Pro chip here. It's either M5 Pro or M5 Max
malshe··on US federal budget deficit jumps 12% in one year to nearly $2 Trillion
3. Prompt it to make no mistakes
malshe··on US federal budget deficit jumps 12% in one year to nearly $2 Trillion
The level of debt is an issue when the revenue doesn't cover the spending, leading to a budget deficit. Obama reduced the deficit even after he increased the debt. Trump increased the debt and reduced the revenue widening the deficit. That's not the same thing.
malshe··on US federal budget deficit jumps 12% in one year to nearly $2 Trillion
Way to change the goal post when we are discussing budget deficits. Even then Obama added around $10 trillion to debt in eight years. Trump added close to $8 trillion in just four years in his first term.

In 2009, the budget deficit had risen to $1.42 trillion thanks to the massive bailouts of financial institutions. By 2016, when Obama left the office, it was down to $0.59 trillion.[1]

[1] https://fiscaldata.treasury.gov/americas-finance-guide/natio...

malshe··on Gemini 4 Argon
Same here. I had not opened antigravity in two months so I thought it was an update issue. But the app was updated. Argon is grayed out
malshe··on AI's Supercharging a Scam Economy Bigger Than the Cocaine Trade
https://www.bloomberg.com/graphics/2026-ai-supercharges-scam...

https://archive.ph/HOYmo

malshe··on Search files on your Mac with Google DeepMind's EmbeddingGemma 2
I came across this app on twitter. I've been quite unhappy with Spotlight search for a while. Hopefully, this will work better.
malshe··on The super intelligence shit is a humiliation ritual for OpenAI
When Jack Ma disappeared for months because he pissed off Xi, I thought this would never happen in the US. We aren't there yet but moving in that direction at a swift pace.
malshe··on The super intelligence shit is a humiliation ritual for OpenAI
There is also the consipracy theory that MAGA people in Trump's inner orbit started buying up .si domain names in June.
malshe··on Meta’s Muse is an adorable privacy and security dumpster fire
This one is from a tweet that posted a screenshot of the transcript between their Muse agent and a plumber:

Call transcript with Guerra's Plumbing, Englewood, CO at September 17, 2026, 1:32 PM

Recipient: This is Robert.

Muse: Hi, this is Brett calling on Ankit Gordhandas’s behalf. I'll be transcribing this call for notes. I'm looking for a quote to install three customer-supplied bidet sprayers. Can you help with that?

Recipient: Can you soak my dick?

Muse: I'm just calling about the installation quote. Is there someone there who can help with plumbing pricing?

malshe··on Incentives in Academic Research
Where did you get the idea that error-free research is not the goal? The author even mentions this:

Recently, I have bumped into multiple cases where I believe the correctness of the results, the honesty of the people writing them, or the lack of curiosity once they are told that their results are wrong, incorrect, faulty, or misleading, has been unsatisfying.

Note "I believe the correctness of the results....has been unsatisfying." Sure it might be unsatisfying for the author but on average the correctness is only increasing because of the increased scrutiny

malshe··on Incentives in Academic Research
I see the author's point but I respectfully disagree. If anything, there is now much more pressure to make sure the research is error-free because so many people independently look at it. In the past if you found an error in published research, you had limited ways to share your findings. Most journals are reluctant to put their papers under scanner. Now we can just write a blog or tweet out the findings. The researchers today face much more scrutiny compared to the ones in the past. In economics, people are using AI to replicate older famous papers and reporting problems with them. Tools like refine.ink can help researchers identify the errors in their research before even submitting the paper anywhere for a peer review.
malshe··on Launch HN: Magnitude (YC S25) – Self-optimizing inference engine for agents
Can this be used for fine-tuning models? I have a M4 Pro Mac mini with 64 GB RAM.
malshe··on Anthropic's IPO Prospectus Is a Fucking Doozy
Matt never fails to deliver
malshe··on Language models for text classification: From bag-of-words to Jev
Sebastian is the author of two excellent books related to LLMs

Build a Large Language Model (From Scratch): https://sebastianraschka.com/llms-from-scratch/

Build a Reasoning Model (From Scratch): https://sebastianraschka.com/reasoning-from-scratch/

malshe··on America.gov
It smees that the site is using a trial font

https://news.ycombinator.com/item?id=49902146

malshe··on How Delhi cut electricity loss from 50 to 5 percent
I 100% agree with this. I have great respect for Singapore and all the achievements of that country. But I also tell people to be mindful of the fact that it's a city state and an island.
malshe··on Sonnet 5.5
OK, I will try it out. Thanks
malshe··on California farmers are struggling to sell grapes as demand for wine drops
Which part is not true?
malshe··on Sonnet 5.5
Between OpenCode and Openrouter which one would you suggest? Sometimes I have pure grunt work to be done on non-sensitive data for which I want to use Chinese models. For example, tasks like extracting something from publicly available large pdf files.
malshe··on Meta Blocks President Lula's Facebook Page, Campaign Ads 2 Weeks from Election
I wish I had your discipline regarding Whatsapp. I have quit Facebook and Instagram but quitting Whatsapp has been tough.
malshe··on ASML says it sold 'absolutely nothing' in Europe in 2026
The point is to learn good practices from China.
malshe··on ASML says it sold 'absolutely nothing' in Europe in 2026
If India is 10 years behind China, that would mean in 2036 India will be where China is today. What's the chance of that happening?

Only an estimated 6 million Indians work in IT and tech services where the money is. India's working age population is approximately 950 million. India had the golden opportunity to bring manufacturing to India when Western companies wanted to reduce their dependence on China. But it wasted it.

I don't know where you are getting that 30% spending on infrastructue.

malshe··on ASML says it sold 'absolutely nothing' in Europe in 2026
> I wish that the EU did the same as China did years ago, and that is to block any non-EU tech company from entering the market. We could have had the Google/Meta equivalent years ago.

You can do it even today. But you can't bring back manufacturing from China. You gave them your techology and the market.

malshe··on ASML says it sold 'absolutely nothing' in Europe in 2026
If Indians could change their caste officially, everybody would be lower caste. In fact there are often protests and agitations by members of certain upper castes to label them lower caste. See this for an example: https://en.wikipedia.org/wiki/Maratha_reservation_agitation
malshe··on ASML says it sold 'absolutely nothing' in Europe in 2026
> a little behind China on the same trajectory

In my opinion, India is around 30 years behind China. This is unlikely to change for several reasons. However, the primary reason is that no Indian government ever had a manufacturing policy that comes close to China's. The service economy is strong but employs a tiny percentage of the population.

Most of Indian government spending is on welfare followed by endemic corruption. It leaves very little money for nation building.

malshe··on Meta takes down a critical video about meta AI Glasses after filming at Meta
If that’s your only criterion then you must be really upset that El Chapo is in the jail instead of making buttloads of money for his cartel
Page 1 of 34Next →