Anil DashI've been trying to find a name or descriptor for people who I think are looking at "AI" broadly and soberly, with a genuinely objective perspective and information that's not captured by the big tech companies but also fluent in the technology behind it. ( @simon would be the exemplar here.) What would you call this cohort? Because I think it's sort of a community without a name, which limits its impact.
D'Arcy Norman đ¨đŚ@anildash @simon "conflicted"? https://darcynorman.net/2024/09/30/i-remain-conflicted-over-generative-ai/
Kevin Riggle
Cory Doctorow
Jessamyn
christina b@anildash
@dweb #DWeb, but thatâs not specific enough.
Iâd look at folks around http://mariafarrell.com/ and https://abebabirhane.com/
Ted Han â
éčĺŽ
Casey
Jason Punyon
Kyle Hughes@anildash Itâs tough because the boundary that defines this group is only present because of the existence of the other two groups on either side of the spectrum. In any other domain these are normal technologists.
sayrer
rojun
Dylan
Ross of Ottawa@anildash ML pragmatists
tante@anildash are people ever "genuinely objective"? You probably mean some form of "lack of strong emotions towards AI" but I'd argue that that - given the real impacts and flaws of AI - creates a bit of a false middle ground.
Even though you might not see it as that I would call a lot of them the (neo-)luddites. But I feel like you are looking for something different that is a lot harder to pin down because they can't be structurally critical which kinda limits the target a lot.
I wonder how useful defining that group is though. Is there a neutral position when looking at something actively harming many of the structures defining our world?
JP
Darius Kazemi@jplebreton @tante @anildash It doesn't really have to be about Grok or whatever. I think the cohort Anil is talking about (evidenced by his mention of Simon) is one that I have a foot in myself -- for example, in my hobby time I'm working with fully local voice assistant tech that uses small LLMs at different levels of the stack, all in service of not sending recordings of my house to Amazon or Google.
"Objective" is probably the wrong word though.
@darius @tante @anildash a lot of so-called "moderates" *do* use Grok for shit though, even if they're just "trying it out" or doing benchmarks or whatever, and i am wary of any effort to normalize that by drawing cohort lines around both them and people at the least compromised edges. it is an overton window of sorts.
like i know we've all read the news but this is so deeply repugnant to me i cannot get past it. i guess that makes me too radical to be on Team Reasonable. https://www.politico.com/news/2025/05/06/elon-musk-xai-memphis-gas-turbines-air-pollution-permits-00317582
like i know we've all read the news but this is so deeply repugnant to me i cannot get past it. i guess that makes me too radical to be on Team Reasonable. https://www.politico.com/news/2025/05/06/elon-musk-xai-memphis-gas-turbines-air-pollution-permits-00317582
@darius @jplebreton @tante yeah itâs an inaccurate word, but kind of gets at the semantic struggle Iâm having
Glyph@anildash @darius @jplebreton @tante âthoughtful but not emotionally dysregulated about the âAIâ brandâ?
@darius @anildash @jplebreton @tante I think I have one foot in this cohort about half of the time, I definitely have my bad days though. (I have a lot of sympathy for how some of the atproto folks obviously seem to feel about âblockchainâ as a word)
Jeremy Kahn@glyph @anildash @darius @tante
I think, sometimes, that being genuinely thoughtful about the hype actually pushes the boundaries of my emotional regulation
Hundreds of billions of dollars of investment in a technology that doesn't work for most of its hyped use cases â that sounds pretty dysregulated to me
Phosphenes
Avery 
Three plus or minus five@anildash @simon
Framkly, there is no such thing as a pure "objective" viewpoint without narrowing the question. "Is marmite good" is not a question that can be answered objectively.
You CAN answer a question like: "do current AI coding methods create product of equivalent quality to humans with less effort", if you define all your terms usefully and use only well-conducted studies instead of expert opinion. But even then the answer is unlikely to be purely objective.
PhD9
loombot 3000@anildash I think Simon's coverage of AI is great, valuable, interesting, and must-read... but... I would not call it objective. He comes at it from his perspective as an engineer with deep skills, lots of different projects to manage, and new projects coming up all the time. For him AI provides a clear subjective benefit... but I don't think he's trying to (or even has the responsibility to!) account for e.g. the impact of AI on the average worker.
Sasha Akhavi
Marsh Gardiner đąđ
Bjørn SÌtrevik
Michael
@anildash @simon almost every demand for objectivity that I see online is really a demand to stop complaining about harms that donât affect the asker. Sometimes they seem to think having formed an opinion means youâre not objective, tho it doesnât. There is a problem that many people form opinions and then have a hard time expressing the reasons behind them, but usually those reasons do make sense. Living in that context, I canât tell what group youâre trying to describe.
Joe Crawford@anildash I don't think such people would be in a named club. They're not in a "fan club," they're not "activists against," they're not "gold prospectors," their livelihood is not dependent on a particular level of popularity of the tech.
IOW: Technologists weighing the value of a new tool against the costs of the tool.
We are in an era of loud street teams, influencers, and "true believers" (in an Eric Hoffer sense). Discourse on "is this hammer better than that hammer" has suffered for it.
Nelson Minar
Evan Dower@anildash
Maybe it would help to define the other groups, too? I think there are several relevant axes.
* AI is not capable
* AI is highly capable
* AI is going to disrupt & solve everything: AI boomer
* AI is going to disrupt & ruin everything: AI doomer
* AI is evil because:
* Plagiarism
* Climate impact
* Jobs
I think the axis you're interested in is the capability axis. The people who say AI is moderately capable in certain domains and can be used productively, though it requires non-trivial skills. Do I have that right?
@simon
Maybe it would help to define the other groups, too? I think there are several relevant axes.
* AI is not capable
* AI is highly capable
* AI is going to disrupt & solve everything: AI boomer
* AI is going to disrupt & ruin everything: AI doomer
* AI is evil because:
* Plagiarism
* Climate impact
* Jobs
I think the axis you're interested in is the capability axis. The people who say AI is moderately capable in certain domains and can be used productively, though it requires non-trivial skills. Do I have that right?
@simon
Adrian Cockcroft@anildash @simon Meanwhile, in a couple of hours from scratch using AI I just built a graph database and populated it with a few years of historical data to act as an MCP data source. It wrote code that worked first time using a Neo4j database I havenât used for a decade, with just a few prompts from me saying what I wanted it to do. Itâs more reliable now than it was a month or two ago.
Jan Lehnardt 
@anildash putz or schmuck
David GerardI mean, if you want to say "AI critics who are not Ed Zitron" you could just say that.
When you require active users of LLMs, that's not "objectivity".
More broadly, you're attempting to reverse the burden of proof.
I commend this thread: https://wandering.shop/@xgranade/115274833212185074
Cassandra Granade đłď¸ââ§ď¸ (@[email protected])
Generally, either "con artists" or "marks." https://me.dm/@anildash/115265668682836682
Hidde@anildash I wonder if technological capability is the most interesting category.
A lot of the questions around AI that warrant objective analysis aren't about the tech, they are philosophical (including questions about what intelligence, knowledge, beauty, truth and ethics are)
T. T. Perry
Mullana