LanceJones
LanceJones on Hacker News
Posted by
LanceJones on Tuesday November 03, 2026 @11:07AM
from the show-hn dept.
https://hn-hiring-2019-2026.lancecj.chatgpt.site/
After seeing Andrej's HackerTube post (and given my sons are both junior-mid developers) I decided to point Codex at 8 years of "Who's hiring?" posts and rip through some tokens on semantic analysis of all that content. I started pre-COVID, thru October 2026, and the data was retrieved via HN API data.5 things that stand out:1. AI is now in half the posts. AI or machine learning appeared in 50%...
1 - 10: How accurately was life at General Magic portrayed in the documentary?
What is the evidence that skills files are an effective way to achieve your goals with the latest frontier lab LLMs?
Fallen? The founders and about 30 researchers took a $2.7 billion exit from Google in 2024.
Neal Agarwal's Absurd Trolley Problems inspired me to go a little less absurd, and also to see how frontier and open weight models would choose in high stake and social stake moral dilemmas.
First I had to create a new set of problems, which I did with my "Trolley Game"... where AI tries to predict your choices based on a handful of warm-up questions (77% accuracy right now). Then I ran 14 models through the full set of 20 dilemmas, 200+ times each, asking them for rationale and predictions about what humans would choose along the way.
Reasoning on; reasoning off. Reversal of choices to test for primacy effect.
Conclusion? We (humans) should be careful about how much control we hand over in terms of tool use in the future... because (1) like humans, AI models don't agree on everything, (2) they are not aligned with us in many ways (depending on your POV, of course), and (3) they "see" us as quite predictable creatures.
Posted by
LanceJones on Thursday September 17, 2026 @11:22AM
from the trolleygame.com dept.
So this appears to mean that Hyundai is effectively taking BD's humanoids "off the market" (B2C/B2B markets). And Softbank wants to take a different humanoids stake in OpenAI's plans.
I always wonder how people can tell. For this particular article, was it the thirty-four occurrences of em dashes with spaces on either side? Something else obvious?
Five EU countries have now approved, including Belgium, which is interesting considering Brussels (ECC "capital").
Posted by
LanceJones on Thursday June 11, 2026 @01:16PM
from the show-hn dept.
My 2024 Model 3 Performance regularly sees its EPA rated range.
$1.2M in Canada after provincial and federal luxury sales taxes. For a 5100 pound, sub-300 mile range, mid-performer with 23/24" wheels. All those louvres, ducts, and aerodynamics for a terribly inefficient EV. Disappointing. (edited because i had $1.1M as the final price)
Just 280+ mile EPA range on a 122 kwH battery. 5100 pounds. 2.5s to 60. Not insane by any standard, ICE or EV.
The idea that 100s of global pension funds don't do their due diligence when investing 100s of millions or billions of their members' future retirement funds is extremely naive. With sincerity, I hope you can find a way not to be so emotional about what Musk says and be more grounded in what his companies and their employees are doing.
Some fair points, but Model Y was the #1 selling SUV in China in Feb and Mar, and 2nd place vehicle was almost half the price.
You think $500K was stolen? Hmm, doesn't sound hyperbolic at all...
I really used to enjoy Fred's writing on Electrek. But after the Roadster referral debacle, Fred's tune changed and has gotten more negative... so negative now that you cannot find any positive articles about Musk or Tesla (some articles you might classify as 'neutral'). If you look at comment counts, it's clear Fred is playing to the "I hate Musk" crowd now for clicks (comments and clicks have similar outcomes for Google Adsense). Electrek's other articles' comment counts pale in comparison. So the incentive is pretty obvious now. Too bad.
But also more deaths. It's the delta that's important.
$200/mth USD for me. Anthropic's oAuth Max20 subscription via their SDK. I've had several 100M+ token days without even a blip.
There's no need to run local models with OpenClaw. I use Anthropic's oAuth Max20 Plan subscription via their SDK...
It will have simulated gear changes if that helps at all...
It's a 4-door, 4-seater. Sporty?
Assuming this experiment involved isolating the LLM from its training set?
BYD sales in January 2026 are down 30% YoY. Not looking great for them in 2026.
Posted by
LanceJones on Friday January 23, 2026 @12:20AM
from the techcrunch.com dept.
OpenAI could be profitable (easily) if it stopped training new models. Whether they will make that choice or not, who knows.
They have positive gross margin and net margin. That's not ZEV credits.
Yes, all messed up for me -- and 0ms deployment (incorrect).
And yet CR says Tesla is #10 (#1 across all EV makers) now in terms of brand and reliability. Fred Lambert has an axe to grind. Why can't HN see that?
Perhaps. I suppose the biggest in history then? $1.4T valuation and 60% of shares held by non-meme institutions (like pension funds, S&P tracking ETFs, etc) when you factor out insiders.