HNHacker News
TopNewBestAskShowJobs

bgwalter

1,468 karma · joined April 5, 2025

submissionscomments
bgwalter··on Rob Pike got spammed with an AI slop "act of kindness"
"Simon Willison REACTS to Rob Pike's unfiltered opinion on AI". We must have the proper spin.
bgwalter··on Rob Pike goes nuclear over GenAI
This is really getting desperate. Markov chains were fun in those days. You might as well say that anyone who ever wrote an IRC bot is not allowed to criticize current day "AI".
bgwalter··on Rob Pike goes nuclear over GenAI
Of course, give people Soma so that they do not revolt and only write meek notes of protest. Otherwise they might take some action.

The Greek philosophers were much more outspoken than we are now.

bgwalter··on Rob Pike goes nuclear over GenAI
Guilt-tripping people into providing more fodder for the machine. That is really something else.

I'm not surprised that you don't understand ethics.

bgwalter··on Rob Pike goes nuclear over GenAI
There is a difference between providing a useful service (web search for example) and running slop generators for modified TikTok clips, code theft and Internet propaganda.

If he is currently at Google: congratulations on this principled stance, he deserves a lot of respect.

bgwalter··on Memory Safety
The annual report of this org is pretty underspecified. There are numerous directors. Are they getting paid under "ops & admin" (11%)? Is "advancement" (12.3%) marketing?

Prossimo gets 6.8%. Does that money go to programmers? Are people whose works are being rewritten and plagiarized offered money to do the rewrite themselves or does it go to friends and family?

Why does an org that takes donations not produce a proper report?

bgwalter··on Who Watches the Waymos? I do [video]
I'm sure that an insurance company would reject larger premiums and make itself obsolete, assuming that claims you make are true.

Hint: Normies outside the tech world are not hell bent on making themselves unemployed. That phenomenon only exists among software engineers.

bgwalter··on Python 3.15’s interpreter for Windows x86-64 should hopefully be 15% faster
MSVC mostly generates slower code than gcc/clang, so maybe this trick reduces the gap.
bgwalter··on Who Watches the Waymos? I do [video]
I would make a movie about UBI recipients wearing shock collars that are supervised by Optimus robots and pluck almonds or apples.

Perhaps we need a euphemism for UBI: Let's call it "level-1 rich".

bgwalter··on Who Watches the Waymos? I do [video]
There are also neighborhood self-help groups who try to "stack" the Waymos into alleys so that they can sleep:

https://edition.cnn.com/2025/06/25/us/santa-monica-waymo-bat...

bgwalter··on Nvidia to buy assets from Groq for $20B cash
They want the Trump Jr. and Chamath connections to be forgotten by the time the media and YouTubers get active again after New Year. It's a standard media tactic.
bgwalter··on Nvidia to buy assets from Groq for $20B cash
Trump Jr. entered at $7 billion. In the meantime Nvidia got permission to sell GPUs to China.

All-In pundit Palihapitiya is invested in Groq as well. It is going well for friends of David Sacks.

bgwalter··on Beijing is enforcing tough rules to ensure chatbots don’t misbehave
China has been more cautious the whole year. Xi has warned of an "AI" bubble, and "AI" was locked down during exam periods.

More censorship and alignment will have the positive side effect that Western elites get jealous and also want to lock down chatbots. Which will then get so bad that no one is going to use them (great!).

The current propaganda production is amazing. Half of Musk's retweets seem Grok generated tweets under different account names. Since most of responses to Musk are bots, too, it is hard to know what the public thinks of it.

bgwalter··on Lessons from the PG&E outage
A human driver does not have to call a remote operator if the traffic lights are off.

This situation does not require a sophistic argument that we are not autonomous because we rely on the sun. If a child walks alone to school without asking for directions, it walks autonomously. If it has to call its parents or uses a GPS phone, it is not autonomous. This is really not that hard.

bgwalter··on Lessons from the PG&E outage
That is a sad state of affairs, I hope it does not make it into the printed edition.

The same applies to "autonomous drones", which are the most remote assisted machines imaginable.

But of course the advertising departments want to evoke an image of the Marlboro man saddling his horse rather than a GPS tracked, surveillance riddled, face scanning, remote assisted contraption.

bgwalter··on Lessons from the PG&E outage
Definition in the Oxford dictionary: "Of, pertaining to, or characterized by autonomy; self-governing, independent; free of external influence or control."

Self-driving car advertisers like Musk or Waymo just want to co-opt this term because it sounds cool. The term also deliberately hides the fact that these vehicles surveil and track you.

EDIT: It is the full definition in the printed Shorter Oxford English Dictionary (which is a large two volume publication). It is understandable that morons downvote it.

bgwalter··on AI Can Write Your Code. It Can't Do Your Job
If you think you are right, shouldn't you tell them to wake up? They would be the ones promoting their irrelevance.
bgwalter··on Perfect Software – Software for an Audience of One
Before LLMs, “perfect software” was largely a myth.

We know, your blog is the first static webpage in existence.

Third, it brings back autonomy.

This is the new talking point. Musk claims that cars that are always connected provide "autonomy", vibe coders claim that the stolen code distributed by Anthropic provides "autonomy".

War is peace, freedom is slavery.

bgwalter··on AI Can Write Your Code. It Can't Do Your Job
So far there are mainly horrible "AI-coded" websites that look like they were produced by the Bootstrap framework in 2014. They use 100% CPU in Firefox despite having no useful functionality.

It is the Mc Donalds version of programming, except that Mc Donalds does not steal the food they serve.

bgwalter··on The U.S. Is Funding Fewer Grants in Every Area of Science and Medicine
If you need links for the "substance" part. Investments of Sacks:

https://techcrunch.com/2019/09/19/cannabis-logistics-startup...

https://sfstandard.com/2024/06/13/telemedicine-adderall-vyva...

Trump relaxes cannabis classification:

https://www.theguardian.com/us-news/2025/dec/18/trump-cannab...

bgwalter··on Larry Ellison provides personal guarantee for Paramount takeover of Warner Bros
He already uses $82 billion in stock as collateral for his borrowing:

https://www.cnbc.com/2025/09/18/larry-ellison-365-billion-fo...

This is another $40 billion guarantee. Imagine he gets a margin call if ORCL drops another 30%.

bgwalter··on The U.S. Is Funding Fewer Grants in Every Area of Science and Medicine
Everything is about "AI", "crypto" and substance grifting. There is no place for real science or useful economic activities like building houses.

Here is the latest fake poll that the Crypto/AI/Substance czar posted and that was retweeted by Musk, who claimed to be an "AI" skeptic not so long ago:

https://xcancel.com/DavidSacks/status/2003141873049952684#m

Getting favors for billionaires is all that these people are concerned about.

bgwalter··on A year of vibes
It is nice that he speaks about some of the downsides as well.

In many respects 2025 was a lost year for programming. People speak about tools, setups and prompts instead of algorithms, applications and architecture.

People who are not convinced are forced to speak against the new bureaucratic madness in the same way that they are forced to speak against EU ChatControl.

I think 2025 was less productive, certainly for open source, except that enthusiasts now pay the Anthropic tax (to use the term that was previously used for Windows being preinstalled on machines).

bgwalter··on Reflections on AI at the End of 2025
I can't help you if you repeatedly misinterpret me. Once you made the first response in this subthread, 4 or 5 of my comments went from 1 to 0 or -1. Cum hoc ergo propter hoc? Maybe.

I'll design a system for the senate that enables outside voters to first turn down the microphone's volume of a speaker if he says that another senator works for company X and then removes him from the floor. That'll be a great success for democracy and "intellectual curiosity", which is also in the guidelines.

bgwalter··on Reflections on AI at the End of 2025
Ah, so you just went through my history and downvoted everything in sight! Thanks for confirming.
bgwalter··on AI will make our children stupid
The authors' (there are two) position is that yes, people are already stupid enough, but it will get much worse:

"We may soon look back on this era of TikTok, Love Island and Zack Polanski as an age of dignity and restraint."

bgwalter··on Reflections on AI at the End of 2025
> Much more important also is that LLMs don't try to scam you, don't try to fool you, don't look out for their own interests.

They follow their corporations instead. Just look at the status-quoism of the free "Google AI" and the constant changes in Grok, where xAI is increasingly locking down Grok, perhaps to stay in line with EU regulations. But Grok is also increasingly pro-billionaire.

Copilot was completely locked down on anything political before the 2024 election.

They all scam you according to their training and system prompts. Have you seen the minute change in the system prompt that led to MechaHitler?

bgwalter··on LLM Year in Review
Karpathy uses the term for all of this in the exuberant paragraph 5. of his blog post.
bgwalter··on Reflections on AI at the End of 2025
> Here we go again.

Indeed, he said the same as a reflection on 2024 models:

https://news.ycombinator.com/item?id=42561151

It is always the fault of the "luser" who is not using and paying for the latest model.

bgwalter··on Reflections on AI at the End of 2025
Regarding the stochastic parrots:

It is easy to see that LLMs exclusively parrot by asking them about current political topics [1], because they cannot plagiarize settled history from Wikipedia and Britannica.

But of course there also is the equivalence between LLMs and Markov chains. As far as I can see, it does not rely on absurd equivalences like encoding all possible output states in an infinite Markov chain:

https://arxiv.org/abs/2410.02724

Then there is stochastic parrot research:

https://arxiv.org/abs/2502.08946

"The stochastic parrot phenomenon is present in LLMs, as they fail on our grid task but can describe and recognize the same concepts well in natural language."

As said above, this is obvious to anyone who has interacted with LLMs. Most researchers know what is expected of them if they want to get funding and will not research the obvious too deeply.

[1] They have Internet access of course.

← PreviousPage 2 of 31Next →