HackerLangs
TopNewTrendsCommentsPastAskShowJobs

AndrewKemendo

23,510 karmajoined 14 yıl önce
https://givedirection.com

https://kemendo.com

Submissions

Slavoj ŽIžek: Should We Grasp AI Not Only as Substance but Also as Subject?

youtube.com
5 points·by AndrewKemendo·12 gün önce·0 comments

So you want to build your own frontier model

givedirection.com
3 points·by AndrewKemendo·18 gün önce·0 comments

Stadium workers have a tentative deal, averting strike ahead of World Cup

apnews.com
5 points·by AndrewKemendo·geçen ay·0 comments

Good Enough to Ship

givedirection.com
2 points·by AndrewKemendo·geçen ay·0 comments

GTA 6 Developers Unionize

rockstarintel.com
758 points·by AndrewKemendo·geçen ay·537 comments

An example of functional slop code

manemasters.vip
2 points·by AndrewKemendo·2 ay önce·3 comments

Show HN: Direction – a 4-week course for people afraid of shipping AI slop

givedirection.com
3 points·by AndrewKemendo·3 ay önce·0 comments

OpenAI on Surveillance and Autonomous Killings: You're Going to Have to Trust Us

theintercept.com
1 points·by AndrewKemendo·4 ay önce·0 comments

"Cancel ChatGPT" movement goes mainstream after OpenAI closes deal with U.S. Dow

windowscentral.com
160 points·by AndrewKemendo·4 ay önce·47 comments

Ask HN: Suggestions for General Tech forums without AI anxiety

2 points·by AndrewKemendo·5 ay önce·0 comments

[untitled]

1 points·by AndrewKemendo·6 ay önce·0 comments

Aldrich Ames, most damaging CIA traitor in agency history, dies at 84

washingtonpost.com
12 points·by AndrewKemendo·6 ay önce·4 comments

Humanoid Teleop with Full Body Tracking Using the Meta Quest 3 and IsaacSim

simintel.co
1 points·by AndrewKemendo·6 ay önce·0 comments

PDFChat: A Local PDF Q&A Chatbot

github.com
3 points·by AndrewKemendo·6 ay önce·2 comments

T-800: 6 foot tall humanoid combat robot [video]

youtube.com
4 points·by AndrewKemendo·7 ay önce·0 comments

Human behavior isn't coherent enough to be a benchmark for AI

kemendo.com
3 points·by AndrewKemendo·8 ay önce·0 comments

Ask HN: Is anyone building an LLM based digital surrogate?

1 points·by AndrewKemendo·8 ay önce·0 comments

The Only AI Explainer You'll Ever Need

kemendo.com
2 points·by AndrewKemendo·8 ay önce·0 comments

Thomas Metzigner on Self Modeling (2016)

youtube.com
1 points·by AndrewKemendo·8 ay önce·0 comments

[untitled]

1 points·by AndrewKemendo·9 ay önce·0 comments

comments

AndrewKemendo
·11 saat önce·discuss
> If the open source community can come up with something like this, just imagine what governments are capable of.

Since ~2022 and accelerated by the Russian aggression against Ukraine, governments are now behind both private and open source for frontier technology.

The companies that captured government contracts in the last century can’t move fast enough to bring tech into the government and national technology policy and funding is collapsing compared to the private sector

That’s new in history
AndrewKemendo
·7 gün önce·discuss
Wow the thing that stood out to me was the number of Darien Gap crossing not being zero.

I was under the impression that was the most dangerous part of the planet for every possible reason.

The fact that it’s half a million a year is crazy!

Something must have changed in infrastructure for that to even be possible right?
AndrewKemendo
·7 gün önce·discuss
I’m not sure I ever met a “DevRel” or similar person that didn’t eventually just come out as the “good cop” sales rep while the “bad cop” sales rep is negotiating with the CFO.

Even the ones who didn’t actively push whatever company they worked for like a commission based sales person, the only reason they are in the room in the first place is to support whatever company technology they work for. I always found it to be a bit gross of a job.

Anyone have a different experience?
AndrewKemendo
·7 gün önce·discuss
Correct

For the set of [all tasks humans can do] there is a subset of [things automation can do at equvalent or better error rates].

And that’s the baseline the baseline is not some platonic ideal that is never reached the baseline for a business operator is cost per delivery /error.

The idea that existing human systems are optimized or otherwise “not broken”, is the key fallacy that people keep making.
AndrewKemendo
·7 gün önce·discuss
[dead]
AndrewKemendo
·8 gün önce·discuss
That is the entire industry of business consulting.

Boston consulting group Bain and MacKenzie make billions of years completely making shit up. same thing with Ernst and young and any of these organizations that make these “future of (insert market)” reports
AndrewKemendo
·8 gün önce·discuss
What’s your definition of AGI?
AndrewKemendo
·8 gün önce·discuss
The entire purpose of automation is to remove a capacity limited human from a continuous workflow because the workflow is more capably achieved with fewer errors than the human

See: traffic lights
AndrewKemendo
·8 gün önce·discuss
I’m unaware of any humans that don’t have this error method also.
AndrewKemendo
·8 gün önce·discuss
Thank you very much for creating this!

I’m gonna send this to people when they tell me that nothing‘s happening and none of this is real

I love this so much thank you
AndrewKemendo
·8 gün önce·discuss
I’ve been peripheral to these systems and yes they are pervasive even in “mom and pop.”

Increasingly they are pushed for insurance purposes to automate “loss prevention” and make it auditable and also help build cases.

If the question is how do you get away from surveillance the answer is “you don’t anymore” unfortunately.

At this point it is pervasive and there is no way to avoid it. I’ve been extremely close to surveillance systems my whole career and it’s to the point where if somebody wants to completely surveil you 24/7 they can do it very easily for very little money
AndrewKemendo
·9 gün önce·discuss
Isn’t that the default though?

Maybe I’m rare in that what you describe is literally how I’ve always done it, but are there more people getting groceries delivered than shopping in store now?

Stores seem plenty full to me.
AndrewKemendo
·9 gün önce·discuss
That’s because we’re almost to the Technological Singularity

Kurzweil puts it between 2029-2032 and that seems right to me

https://en.wikipedia.org/wiki/Technological_singularity
AndrewKemendo
·10 gün önce·discuss
Thanks yeah exactly!
AndrewKemendo
·10 gün önce·discuss
Were previously open to cold emails
AndrewKemendo
·10 gün önce·discuss
So……profitability

Is the person defending Anthropic in the room with us?
AndrewKemendo
·10 gün önce·discuss
> The author is pretty obvious and exhaustive about what he means by "losing": AI capex bubble is unsustainable, AI revenue is circular, no meaningful AI compute demand outside of OpenAI and Anthropic, AI products are mediocre at best (and still heavily subsidized, at that), AI is causing various mental health crises, OpenAI lost $20.9 billion on $13 billion in revenue in 2025.

So then it’s just profitability modulo “various mental health crises”
AndrewKemendo
·10 gün önce·discuss
Losing at what?

The author never states what would constitute winning
AndrewKemendo
·11 gün önce·discuss
> Well... again, I'm giving the LLM total freedom to destroy what's there and start from scratch

Why would you use it like that as a default?

Like the first thing you should do when talking to an LLM about a project is to have it discuss exit criteria, scoping, rules, and managing the smallest thing to ship to get iteration started
AndrewKemendo
·12 gün önce·discuss
Thanks! I appreciate it