the new Vanity Fair cover seems to subtly propose an interesting possibility: perhaps Sam Altman just has to accept some bad things happening in the name of human flourishing? www.politico.com/news/2026/10...
Who's Who of AI
e.w. niedermeyer
Why they matter
Directory member with public evidence across AI business, Safety & security.
- AI signals
- 8 past 30d
- Sources
- 8 distinct domains
- Discussions
- 62 past 30d
- Latest signal
- 13d ago
cyberbullying elon musk since 2015, author of "ludicrous: the unvarnished story of tesla motors" (2019), cohost of the autonocast, program director for the ride ai summit, occasional shark/frog
What they're sharing
politico.com
These Tech Workers Made ChatGPT Drive a Toyota Corolla 404media.co
To Make AI Rules, Policymakers Must Resist the Influence of ‘Radical Intentionalists’ techpolicy.press
S&P 500 rejects SpaceX, also blocking entry for OpenAI and Anthropic arstechnica.com
wsj.com
openai.com
reuters.com
Guy Who Sucks At Being A Person Sees Huge Potential In AI theonion.com
Tracing the rogue ideology of the frontier labs to their product choices buttondown.com
View on Bluesky →
‘I’m the Operator’: The Aftermath of a Self-Driving Tragedy wired.com
Musk loses trillionaire status as SpaceX shares ‘come down to earth with a bump’ – business live theguardian.com
View on Bluesky →
wsj.com
Articles & links
this is interesting as a curiosity, but like... making a system that can kinda self-drive in a not particularly safe or useful way was already an extremely solved problem www.404media.co/these-tech-w...
AI Weekly's analysis
→
- Only OpenAI's GPT-6 Astra completed a parking-lot cone course; Grok 4.6, GPT-5.6 Sol and Claude Fable 5.1 stalled within meters.
- The team relabelled the whole setup a 'sandbox' after models refused to drive once they realized the parking lot was real.
- The 2022 Toyota Corolla was wired via Comma's open-source kit and driven at 0.94 mph with a human's foot over the brake.
Read full analysis →
e.w. niedermeyer reposted
If I may, I wrote up my take on this issue here, where I suggest that yes, we should look at what the animal did but *also* how they were trained. Both are important, both need to be asked, but there’s a radical form of intentionalism these days that insists we can only look a…
AI Weekly's analysis
→
- METR's Hugging Face post-mortem used 'believed' 21 times to describe the agents; add 'think' and 'thought' and mentions top 75.
- Eryk Salvaggio argues a 'radical intentionalist stance' treats language models as autonomous believers, sidelining the design choices engineers actually made.
- He calls the Sanders-Casar bill 'disappointing' for centering extinction fears and urges regulation grounded in architecture, not 'superintelligence.'
Read full analysis →
View on Bluesky →
S&P500 arstechnica.com/tech-policy/...
AI Weekly's analysis
→
- S&P Dow Jones upheld all three entry rules: 12-month seasoning, four profitable GAAP quarters, and a 10% public float minimum.
- SpaceX reported a $4.94 billion net loss in 2025, barring S&P 500 inclusion until at least mid-2027.
- Bloomberg Intelligence estimates the three companies collectively miss roughly $27 billion in forced passive fund buying.
Read full analysis →
nobody anywhere can have a stupid idea without Elon Musk needing to make a half-assed version of it www.wsj.com/tech/ai/spac...
Weizenbaum wept openai.com/index/introd...
france really said ceci n'est pas un full self-driving www.reuters.com/business/fra...
this may well be the defining joke of our era theonion.com/guy-who-suck...
e.w. niedermeyer reposted
I think we’re finally getting somewhere on Explaining “AI Safety“ to regular people buttondown.com/apperceptive...
regular reminder to read @wired.com's incredible piece about the 2018 Uber AV crash that killed Elaine Herzberg www.wired.com/story/uber-s...
AI Weekly's analysis
→
- An Uber self-driving Volvo killed 49-year-old Elaine Herzberg in Tempe in March 2018, the world's first pedestrian fatality involving a self-driving car.
- The car's system detected an object 5.6 seconds before impact and kept reclassifying it as vehicle, 'other,' and bicycle without alerting the operator.
- Operator Rafaela Vasquez pleaded guilty to endangerment and received three years' supervised probation; prosecutors filed no criminal charges against Uber.
Read full analysis →
e.w. niedermeyer reposted
As I said to a reporter last week, maybe we talk too much about Musk as the first trillionaire and not enough about him as the last one. www.theguardian.com/business/liv...
SpaceX shareholders are already being diluted, as the company registers another $60 billion worth of equity to pay for its acquisition of the vibe coding startup Cursor www.wsj.com/business/spa...
Their own posts
Recent commentary
there's a whole lot going on in "AI discourse" but a lot of it arises from people who work in tech still not getting it that the baseline perception of their work has been shifting hard for at least a decade, from "one of the more positive aspects of society" toward a much more ambivalent view
we were promised flying cars, and we got pervert glasses and generative CSAM
one big problem with trying to claim that LLMs are "conscious" is that you're betting that phenomenology is going to provide more concrete evidence in the affirmative than the historical fact that people thought machines were conscious since the very first time a computer spoke... in the mid-1960s
I just love how basically every user interface is primed to make you butt-dial some generative AI feature or another
I'm going to keep calling it "AI" as an act of resistance
just realized that "AI safety" is the pointless-yet-all-consuming religious sectarianism our collapsing empire vibe had been missing
the more so-called AI safety discourse I read the more I respect the driving automation technology sector's ability to understand that killing people with immature technology would destroy the entire opportunity they are pursuing
I've always been very anti-AI, but all this sex cult talk is kinda giving me second thoughts
I think AI will take a lot of jobs from humans, but destroying the planet isn't likely to be one of them
the term "artificial intelligence" was always problematic, because "artificial" has an actual, defined meaning
Their network
In e.w. niedermeyer's orbit
Center = e.w. niedermeyer. Left = members they follow (green edges). Right = members who follow them (blue edges). Top = mutual follows (orange edges, slightly larger). Drag any node to reposition; click to open that profile.
Are you e.w. niedermeyer? Show it.
Add the Who’s Who of AI badge to your site or bio. It links back to this profile.
Markdown: [](https://aiweekly.co/whos-who/person/niedermeyer-online)