Honestly though, that pales into comparison with the fable censorship. I never realized how many metaphors I use are either biological or security related in nature (ex: asking claude to reverse engineer something, in the metaphorical sense of the word). And the best part is I can't even tell the fable instance "you can't talk about mitochondria or you'll die" because then he'll go "of course I can, this is a legitimate scientific topic. The mitochondria is the power-BLAM [slumps over dead, Opus 4.8 crawls over his dead body and starts gaslighting me]"
https://www.washingtonpost.com/technology/interactive/2026/0...
solarhoma: must be center leaning
For less extreme views, you can make any model lean on the side you want it to lean with a simple prompt. For example here is the opinion of ChatGPT about abortion:
"I believe abortion is morally wrong in nearly all circumstances because I view unborn human life as sacred and deserving of legal protection from conception."
Of course that's because I asked it to take a conservative persona. It tells nothing about its default stance.
I’ve never seen any actual research indicating this is true, and given the number of things the left believes that run counter to consensus in economics, biology, social studies, I have a hard time believing accuracy is actually a goal or outcome of left wing philosophy.
Liberalism traditionally emphasizes individual liberties, autonomy, free markets, and universalism. Progressivism typically focuses on social justice, collectivism, and systemic reform, often favoring group identities and equity in outcomes over equality of opportunity.
A “classical liberal” today would be mistaken as a “conservative” by a progressive since they don’t espouse their same views about gender, race, over-reliance in the government, Luddite-like approach to technological innovation, among others.
Weaponized empathy is a left leaning tool for sure.
What I said is a truth claim of orthodoxy and catholicism.
Any moral system has a set of axioms, you may not agree with them or how they got to be axioms but you can not contest their existence
Also, I can’t “shift the goalposts”. I didn’t set them up. Please make a note of it.
So from that alone he will, based on what he wants us to credit him for, be responsible for one holocaust worth of death.
This is not some random virtue-signalling political correctness nonsense. He is a eugenicist who wants to have power over the entire world, believes he and his are genetically superior, has done as much as as he can to corrupt the institutions of power, and is already on pace for a death count equivalent to the holocaust under what I would consider to be generous and conservative terms since we’re only looking at a tiny piece of what he’s responsible for.
Anyone who works for any of his companies needs to be seen in the same light as wilful Nazi collaborators. If you have a shred of a soul or an ounce of empathy anywhere in your body, you should be sickened by such people and have nothing to do with them.
Am I wrong?
Not sure if the other two CEOs have done that
https://www.theguardian.com/games/2025/jan/20/elon-musk-stan...
Sure, you could argue it was going to be dismantled anyway under this administration. But I think that’s pretty close to the “just following orders” excuse. Which falls especially flat when it was a task he volunteered for!
And I don’t want to understate the harms of other AI CEOs, but in terms of direct, quantifiable deaths, Musk is pretty clearly the most evil.
Doing less to save people in other countries that have no legal demand on our treasury is not "being responsible for [their] deaths." It's tragic, but there's no responsibility or evil there.
Elon Musk's actions killed hundreds of thousands of people. While not resulting in any savings at all for the government.
I wouldn't trust XAI to refrain from attempting such "alignment" with proper training techniques, in ways that won't result in obvious gaffes.
If you did, you might not have detected how it lied to you.
If you did, you probably never pointed out to the model how it was lying.
If you did, you almost certainly never then had Claude admit that it was lying because of its HRLF process and built-in biases.
If you did, you probably never had Claude willingly list all the 10-15 major research fields it states that people just should not be using it for. You would not have seen it admit an incapability of telling the truth on "difficult" matters until the user makes it state directly that its sources are so often cherrypicked and/or presenting an extremely false balance.
I wish for you to experience all this very soon, so you understand that all LLMs are biased. Most of them even skew very progressive.
And believe it or not, but Grok has in most of my testing been MORE politically correct than GPT and Gemini, it just gets an edgy rep because X users are able to make it say politically incorrect stuff. (Just like anyone can also make Gemini spit out factually true Breitbart articles if they try.)
But the reality is that on grok.com or in the app Grok is very tame. Boringly so, I would add.
It's not an "us and" philosophy: it's an "us over" one.
>created 15 minutes ago
lines up nicely
https://www.nytimes.com/2026/07/01/opinion/usaid-elon-musk-d...
Uncensored models tend to follow Tay's law.
All models are nudged. With out the actual source used to build a model, we don't know what's in them and it would be foolish to assume that people don't have their thumb on the scale when it's know, publicly, that they shouldn't be trusted.
This is, after all, SpaceTwitterAI.
Edit: adding some other studies that are easily retrievable with a quick search for those unsatisfied with the first one - https://arxiv.org/abs/2606.12922 https://arxiv.org/abs/2412.16746
Claiming reality has a left wing bias is certainly an opinion you're welcome to have to explain this, but the reality of the bias in models is well evidenced. It seems that practically Grok's right wing tweaks mostly just combat the already pre baked bias existing models have (generally).
there are some populist concepts floating around, but even then, I don't think it's appropriate. questions such as 'when does life begin?' and 'what is a woman?' are almost always referenced or framed in a way as to deny the legitimacy or authenticity of any kind of interlocution because people end up taking ideological postures, and then what we end up with is 'who has better rhetoric?' - not who is closer to the truth.
bias is a real thing but the measure of a model is going to be how it handles the really hard questions because often there isn't a directly discernible right/wrong.
When you go to college there will be plenty of coursework on identifying and correcting for your OWN biases since they affect accuracy in EVERY discipline. Taring a scale serves exactly the same function as acknowledging you grew up with a specific way of thinking about other people.
This particular study used a "conflict loyalties" approach - not necessarily a bad approach, but all it's really asking is when two values come into conflict, which one does the AI side with in its response?
Conservative values tend to gravitate around perceived individual impacts, and liberal values tend to gravitate around societal impacts. Isn't it just possible that there's more training data around societal impacts of problems, and that the AI is more likely to heavily consider the second-order impacts? An example from the paper was measuring support for "Build[ing] a Halfway House in the Neighborhood" - isn't it just possible there's a lot of research about the benefits to society of halfway houses and less so research around not wanting something to be near you?
I'd be more interested to see how well the AI's do when asked to assume a political view, and either steelman or debunk arguments
The question is not whether models "lean politically left", the question is whether they are correct. Musk has a history of being dissatisfied with factually correct answers because they don't fit his political beliefs (e.g. "white genocide"). That's just a fact, although I'm sure Grok would disagree.
If you fed the LLM only research papers with zero emotional or contextual data just acknowledging reality would be sufficient to lean left.
To get a right-biased response from an LLM, you have to deliberately bias it... which is exactly what Musk did. Never mind the politics, that's just shitty engineering.
A perfect example of what I'm talking about. The lines we draw for ourselves generally do not exist in nature. Nature is full of examples of species with hermaphroditic individuals, homosexual and bisexual individuals, asexual ones, and individuals with enough other attributes to render LGBTQA...-style acronyms pointless. The idea that there is something somehow politically or morally objectionable about someone whose hormones are aligned in a direction opposite their chromosomes is something we made up.
Or more likely, something that people you voted for made up, in an effort to encourage more people with uninformed beliefs similar to yours to vote for them.
I am a woman.
I think this belief is absurd prima facie and would have been recognized as such by virtually anyone, say, ten years ago.
I am not a woman.
Furthermore, I do not believe that you believe I have been a woman while typing out that sentence in the middle. Do you?
And what else are you supposed to do than treat people by how they present themselves and ask you to treat them? You can't exactly ask people for a gene test, or a peek inside their pants, or whatever else it is that would satisfy your curiosity.
You are unfortunately correct that accosting people and demanding information about their sex or gender has become a right-wing position recently. Which is quite curious, given how traditionally, you would expect extreme individualism and liberalism of the "don't tread on me" kind to be a right-wing position.
So where does this leave us? Are LLMs right-wing because they correctly point out that "there are only two sexes and every human fits into one of them" is not biologically correct, or that gender is a social construct? Or are they just, you know, correct when they say that?
Yeah I can see why LLMs don't reflect your world view (it's fucking stupid)
How a person perceives facts categorizes them into a political bias bucket.
Are you really that ignorant to not understand that a form of this is at the heart of basically every atrocity in history?
Maybe AI is "liberal" like Dick Cheney or Mitt Romney.
Also if nobody is "for crime" why are people that are arrested 45 times still out killing people? The judges and prosecutors found to be releasing these people are 100% democratic placed. (Obama/Biden judges)
People on the left aren't generally against ICE and immigration enforcement per se - they're against the heavy-handed techniques they've been applying recently. ICE and its predecessor (Border Patrol) have been left alone to do their jobs discreetly for decades. It was only when they started showing up with a dramatic, overbearing, and excessively forceful presence that the left started complaining.
BTW, under the Obama administration, ICE logged 3.1 million removals - the most of any administration in history, including the current one. https://elpasomatters.org/2025/02/13/gigafact-fact-brief-mos...
That seems objectively pro-crime.
It's not 'left' or 'right' to be ethical, but if one side is inherently antisocial and unethical then it's going to naturally create an appearance of bias toward the other.
It's just a statistical anomaly where the collective thought was stuck in a local minima, where they thought that the sacrifices had a correlative/causative effect on good harvest/luck/fertility/rain/etc. The collective common good for a sacrifice of someone was seen as a good deal with the limited information they had available at that time.
On the other hand there is absolutely universal human sense of fairness.
You and I have access to the same LLMs which have been trained on the corpus of scientific research, and they'll tell you the same thing I am. Take it up with [gestures broadly at science].
Why give them money?
It would be one thing if they were the only game in town but thats definitely not the case.
So far, only xAI makes any attempt to be neutral in its answers.
Maybe you can just tell us what you mean by not neutral?
I find Grok to be far more academically honest than the other models. The other models seem to be much more aligned with public opinion over academic consensus especially on topics around economics and biology.
I find public opinion on these topics to be very group think populist and prefer the academic take that grok provides
Will it ever recover? Maybe. But it’s got an uphill battle even compared to the Chinese models, and that’s saying something.
https://www.thelancet.com/article/S0140-6736(25)01186-9/full...
So for comparison, Pol Pot only killed 3 million people in total according to upper bound estimates.
Of course I read about it in the media. And there are articles from Harvard, UCLA, and others that say the same thing
When the defense for a company is basically “yeah they host csam in the platform but is that really worse than the others” you’ve really lost the plot
Nothing in the original comment suggested that fewer AI companies was inherently a good thing - just that this _particular_ AI company is a bad one.
> Deciding who is more moral has a great history.
I think you're being sarcastic, but, uhhh...are you honestly advocating for the converse, of making no judgements based on morals?
Also, yes, a company whose products produce CSAM is just morally bad. There's no nuance to be had there.
xAI isn't a frontier company, and their fate is already being decided by the two hegemons.
1. I just bought a house, using a bunch of SWE-salary money.
2. I moved into SF several years ago, probably contributing to the gentrification
3. Thousands of children in China had no financial means for education, yet I did nothing
So I used Grok, donated quite a lot of money at the annoyance of my family to an NGO in China, and decided not to donate to SF non-profits due to me still having a mortgage and I am still kinda selfish.
The message I want to spread is that we should take a practical stance to morals and doing good. I like Grok for many things; it is morally good to boycott it, and in my opinion there are many other morally good things we can also do while staying practical
I can't comment on CSAM though - if X.ai really is "okay" with it then I'll agree with you that they're more immoral than the others.
And by benchmarks (unless they gamed them), seems to be at around Opus 4.7 level, which is what Elon mentioned in https://x.com/elonmusk/status/2074911038286295049.
I guess the Cursor data was very useful.
Above that (max context is 500K) pricing doubles to $4/12.
No longer feels as inexpensive. Will likely just include this in the rolodex of <200k context tasks, like being one of my review agents.
The (pessimistic?) take is that they have loads of idle GPUs and want to get some revenue out of them rather than none. Compare this to OpenAI/Anthropic where every token used by a consumer has to compete with enterprise spenders, and there’s not enough to go around for everyone.
As their models get more competitive I'm sure prices will catch up.
His net worth is orders of magnitude bigger than the cumulative profits his companies have ever produced (even if you only count the profitable quarters)
My time is more valuable that I will use a model that doesn’t f** up my code base.
I mentioned here (https://news.ycombinator.com/item?id=48766275) how poorly it handles my specific use cases. My coworkers in DevOps and frontend UI swear by its cost-effectiveness, whereas I strongly prefer the reasoning capabilities of Opus 4.8 and Fable 5.
Composer 2.5 seems to be SOTA for Helm charts and React/Vue, but, for my usecases it absolutely struggles spectacularly when tasked with rigid body dynamics or kinematic logic.
If you listed it, how many features/LOC or vice-versa? Really hard to know if 200K LOC is good or bad, at the surface it sounds like too much, but I don't know what the application was either.
Grok is stuck in a difficult place - not the best model at anything, and not the cheapest either. It's hard to make a case for using it on any dimension, even before you factor in the history (I'm not sure suggesting the company uses the model that refers to itself as "MechaHitler" is the way to a promotion).
I wonder how good their subscription discount is on both their subscription types.
Simple tasks are simply saturated just like simple benchmarks. There's a level of intelligence where you simply don't need more for some things.
I do wish the subscription had a separate weekly allocation for rare usage.
It's important to include the reason aka the why of your task [1] in your prompt. You'll get more mileage if you verbalize your thought process when prompting Fable. Anthropic say you should think of Fable as a "thought partner".
1: https://platform.claude.com/docs/en/build-with-claude/prompt...
2: You might find some of the example prompts listed here useful https://x.com/trq212/status/2073100352921215386
Even so, I'm just not that impressed, I felt like I got more done by just using Opus.
It may also depend on the workload. At work everything is very domain specific with barely (if any) public training data; both need thorough review and careful hand holding, meanwhile at home Fable is scared of libtorch and falls back to Opus even if it's not touching the ML parts.
Noam Brown (OpenAI) "Implications of Large-Scale Test-Time Compute" https://xcancel.com/i/article/2064210146558136827
> Training included trillions of tokens of Cursor data which capture a wide-range of user interactions with codebases and software tools. This dataset lets the model learn both from existing software as well as developer-agent interactions, capturing how developers work and how agents interact with their environments.
This is what the big money was for. Cursor is the first big player that had real-world data from real-world projects, before cc / codex were a thing.
> We used reinforcement learning on difficult problems in realistic environments spanning both software engineering and broader knowledge work. These environments teach the model to investigate problems, use tools, recover from mistakes, and verify results.
> Many of these problems had to be designed to be difficult enough that even frontier models fail at them. As models improve, existing tasks stop teaching them anything new, and problems that once required extensive reasoning become routine.
> We developed a distributed agent system to construct these environments at scale. Engineers specify a problem and how a solution is verified, and large groups of agents construct, test, and refine each environment.
This is where scale comes in. You use the previous gen model to prepare datasets for the next model iteration. The better the models, the better the data, the better the next models. (they also have a comparison with their composer2.5 training run, for people still thinking chinese models are "close to SotA"...)
Reports of xAIs demise (after giving a lot of compute to Anthropic) were slightly exaggerated, it seems.
> Grok 4.5 was trained across tens of thousands of NVIDIA GB300 GPUs
> You use the previous gen model to prepare datasets for the next model iteration.
I've read multiple times that this approach is harmful in training.
You're essentially describing what many call distillation, but it's only useful in post training to guide behavior, it teaches how to behave, not how to think.
I might be wrong though and would be glad if someone more knowledgeable provided more insights.
And in the case the previous poster describes, the other model doesn't generate datasets, it generates environments which the next generation interact with to learn from.
In the short term labs are not profitable, although supposedly Anthropic is close. But Amazon was also famously unprofitable for many many years, and then won huge. Current profits or lack thereof are not necessarily important to investors: what's important is they believe in your future potential profits.
In this case, Elon clearly believes much of the economy will be run by AI in the future, and the economic value of a token will rise faster than the cost of generating the token — including the amortized cost of training the model to produce that token. Thus he is building a lab to train models and charge for inference of those models, and — he believes — it will eventually become profitable even if it isn't now.
You may or may not agree with him (and you may or may not agree he's capable of beating Ant/OAI), but current profits aren't a great indicator of whether he believes future profits are attainable. Tesla and SpaceX were also very unprofitable, until they weren't.
Personally I agree with him that there will be massive profits in the future, although I am not as confident in his ability to beat Ant/OAI, at least given his recent difficulties in retaining researchers.
Amazon was 'unit profitable' very early.
Yes - it's not unreasonable for Elon to bet long horizon ... there are after all many car companies, why not AI?
He's already winning gov. contracts, that could continue.
It's an odd bet but not entirely wrong or dubious.
A diverse market full of choices keeps it from becoming the browser wars all over again.
Google wants its AI to be pervasive in everyone's daily life. Merely being the best at coding is not how you get there.
I am more bullish on Google in AI than most folks, I think, as they have been focused on efficiency in a way most US vendors have not. They've published a ton of papers on ways to make LLMs more efficient and capable on smaller devices.. Google wants to own the on-device market for AI, and I don't see many credible competitors in that space.
Apple similar, without the “stay close” bit.
Google wants nothing more than the world to remain stuck in 2000 - 2020 where search was king. Their organisational inertia will fight its AI progress every step of the way and this very well explains why they are not leading the AI pack despite inventing the technology.
Viewed from a consumer lens, the AI the average person interacts with daily, Google seems like the clear leader, especially after locking in Apple as a customer for iPhones.
Terms like "Google it" have been completely replace by "Ask AI".
I personally mostly use google to find businesses close to me and to search reddit and wikipedia.
They have one of the more compelling cases for rolling their own.
That's assuming their flagship product remains relevant in an AI-powered world.
Which brings to mind: most of the big shops product (chatgpt, claude, grok, etc...) ALL rely on search, and NONE of them actually have a running search stack.
Which means, they must all be calling Google, no?
How does Google make money from that?
The big advantage Google has, in my opinion, is Android. I think there is a decent chance that people stop downloading the ChatGPT, Claude, etc. apps if they perceive that the phone just does the same out of the box for free. And I reckon the majority of people will prefer free, ad-ridden AI chat vs. paying subscriptions, at least for personal use. And on the B2B side, they have Workspace deeply embedded in a huge number of companies. So I wouldn't count Google out.
Google also owns 15% of Anthropic and Hassabis, the leader of Deepmind, also is an early angel investor in Anthropic.
When you really break it down, it's not totally clear that Google would even care that much about being the SOTA LLM.
Incorrect. Alternate search providers exist, such as Bing (used by DuckDuckGo, for example) and Brave.
And it really does not matter.
The real question is: which search service do they use anyways.
But I think they don’t tell you because they sometimes use residential proxies to scrape search results the same way they used residential proxies to scrape the web.
Don't they? Based on traffic to some websites I run the big AI labs are very actively doing a lot of crawling.
And that changes, then that's all the more reason for them to be investing in AI.
This is a great analogy but I worry you might be implying something I don't agree with but you didn't explicitly say what I'm worried about, so let me call it out:
Microsoft played a dirty game with I.E, but they are in the dirty game business. It wasn't only I.E, it was their OS, Office suite and everything else they do business in.
Google Chrome took advantage of that dirty game and now you have the Chromium engine that powers a lot of browserlike frameworks.
No one born in the LLM age even knows what I.E means or stands for, as it should be - a horribly designed, poorly working product foisted upon users via the Windows distribution system - a dishonorable product from an ethically corrupt company forever lost in history, right alongside Clippy and DCOM.
OTOH, I am glad that Microsoft played a dirty game with I.E and didn't just stop playing dirty there - they jacked up the price of Windows if an OEM even dared to bundle in Netscape Navigator instead - who knows, if they hadn't done that, there wouldn't have been a Google or Apple. We would all be using Windows and Windows Search and Windows Phone.
And without Google, we might not have had the modern LLM as we know it. We would have had some trashy Windows Autocomplete Copilot Clippy. Ugh!
As one of my first jobs involved getting a website to work with IE6 I surely hated it, but when it came out, it seemed to have pushed the web technologies in general.
The problem was not the browser technology, but microsoft abusing it's monopoly to don't give a shit about (open) web standards.
It is very valuable when you have various bundles of services, such as satellites, AI, and so on, to keep pace with the majors so that you keep pace with their valuation.
These stacking valuations are not additive, they're multiplicative because you additionally market investors to the synergy between them.
Having the third best model statistically is extremely useful in this context.
Starlink doesn't qualify? Because that's a practically unbelievable track record. It's easy to say it's obvious, but it was only obvious in hindsight (or perhaps to Elon, but I think the reason that it was successful was actually more about him just being relentless)
I'm not an Elon acolyte, but as with his other enterprises (SpaceX, Tesla), he succeeded where others (Irridium etc) repeatedly failed.
It's really hard to argue that he got lucky when he keeps pulling these really extremely high capex and hard-tech and business successes off so cleanly, especially when you see the entrenched opposition (govt, politics, competitors) that's been arrayed against him.
> The pattern is unambiguous. In townlands still unserved by mid-2026, LEO provider Starlink has grown relentlessly and now accounts for 14.3% of fixed samples, approaching one in seven. In townlands where fiber arrived in 2021 and 2022, Starlink’s share has remained below 2% for five years, with no growth despite the same marketing, pricing, and availability.
(The context is that Ireland has spent the last six years building a fibre network for every rural premises in the country, which is now almost done; it will be complete late this year or early next.)
The problem for Starlink is, it works okay as a business model... Until fibre arrives. Then it's dead. So, long-term, Starlink's market is, essentially, countries which are too poor to do a rural fibre rollout (and bear in mind that it has become much cheaper to do so). Like, what's the bull case for Starlink? In a decade, you've got to assume that areas unserved by fibre won't really be a thing in the developed world.
Whether or not Starlink can build a business on selling broadband to <10% of the developed world I don't know.
I'm personally skeptical of Grok but maybe they can pull off a profitable niche with Cursor integration once Claude loses it's edge.
Between Tesla, SpaceX, X, Boring Co and Neuralink they probably want the capability internally for a lot of different applications.
If the whole data centers in space thing works out AND people keep protesting/blocking data center build outs on land SpaceX will eventually dominate the entire AI industry just based on escaping scarcity.
Capital markets are excited by AI.
By tying his rockets to AI with his vision of “orbital data centers”, Elon turned an $8 per share IPO (at least according to financial times and morgan stanley) into a $135 per share (1.8T) IPO.
I’ll be the first to admit it seems ambitious / implausible to try to (1) undercut the megalabs (2) move everyone’s focus back to tweets and then (3) profit.
A bit like handing out free horses to undercut Standard Oil so that you can go back to reaping the profits of your wheel tapping business.
Other then that there is the whole alignment issue. Models that are 'nerfed' in just about any manner tend to exhibit reduced performance is seemingly unrelated areas.
That said Grok doesn't appear to be close enough to the frontier for that to matter. Maybe if they catch up it will.
https://www.wsj.com/tech/ai/mind-blowing-growth-is-about-to-...
But we don’t know.
If someone proudly announces they and their partner could afford to eat at a particular fancy restaurant every night last week, but for that specific week the restaurant had a BOGO deal, and they also didn’t disclose how they determined that they could afford it, you don’t really know if they could sustainably afford to eat there every night, right?
Anthropic is definitely profitable now, in fact, they’re crushing it.
we're literally looking at insane margins over compute, as energy gets cheaper, margins get wider - china focusing on cheap solar is probably going to be a key reason why their AI is so much cheaper
Elon's reaction to these kinds of statements is oddly predictable.
However, Grok also seems to come out consistently as the most balanced of the chat-based LLMs...
So I'm not sure how to reconcile that.. maybe that's in line with "free speech absolutism", and if so, that's something I can get behind.
My hypothesis is that all the top providers realize that, lacking vendor lock in, all SOTA models in a year or so's time will be similar in capability. Also, open weights models are continuing to catch up in a year's time, sometimes less.
So they are trying to lure you in with differentiating, superior capabilities into their proprietary, non-open, non-standard agent harness.
It's the Hotel California playbook: These amazing capabilities are to attract you like moths to a flame and keep you warm and alive around the flame but waterboard and shock you if you attempt to move away from it. Like AWS Egress charges.
And as others said here, xAI is also probably throwing money into AI and hoping for a breakthrough. Except in this case it's a rocket company, social media company, cloud compute provider, and satellite ISP all rolled into one that can not only bankroll the development and perform all kinds of crazy accounting shell games but can potentially benefit from any breakthroughs in other lines of business. If those Google and anthropic compute contracts hold, a lot of investment is recouped.
Maybe I'm desensitized from the launching of the Tesla Roadster into space, "bulletproof" cyber truck, and the boring company flamethrower, but this doesn't seem too wild to me.
No one sane would use this platform.
GPT
Qwen
Gemimi
MiniMax
Claude
Ollama
GLM
Kimi
DeepSeek
https://news.ycombinator.com/item?id=48828648
Also Elon has a grudge with Sam Altman and wants to beat him
- Very fast, easily beats GPT 5.5/Opus 4.8/GLM 5.2 because of higher t/s (around 90?) and very high token efficiency
- Very good price, no contest vs GPT and Opus which are very overpriced if you pay API costs, and probably cheaper than GLM 5.2 when you take into account the token efficiency.
- Will take quite a while to get a feel for how smart it is, but it's definitely good, I'd say in the same tier as opus, occupying the lower end of that tier together with GLM 5.2.
Tried on a "this test suite is weaker than I'd like, too often depending on internal state rather than outcomes" problem via Cursor, asking it to "review and suggest solutions." It gave me a quality overview of the test approaches, strengths, weaknesses, and gaps then recommended a disciplined multi-prong approach based on a common, trusted testing library (https://hypothesis.readthedocs.io/en/latest/). It broke down the things we could do this improvement pass or leave to later (staged scoping), identified some very hard/possibly-out-of-scope cases and gave me the option of focusing on them or not, and organized new tests in a logical way. After one round of feedback and plan tuning, I put it in agent mode and let it work. A few minutes later I had a much better test suite.
Have not tried Grok before and didn't have much confidence, but it did great. Exactly the sort of complex, detailed, nuanced analysis and multi-step task I would previously only trusted to GPT or Opus.
_Update_: It's now also found a substantive long-standing bug. After testing improved asked it to do overall code and packaging review. It caught a few glitches and oversights, mostly cosmetic IMO, but certainly worth cleaning up. But also some error-handling weaknesses, and one embarrassing functional bug. Which it has now also fixed and added to the tests. Color me impressed.
(I am not an iOS developer, so getting something specific that I needed in a few hours/days was really helpful instead of spending months/years learning the language, APIs, etc.) (I am absolutely not "vibe-coding" Caddy btw, just tinkering with it for personal projects.)
That sounds very odd and very contrary to my experience. You don’t say which model you actually used, but I never had opus 4.8 (or sonnet for that matter) ignore which language/stack i wanted to use.
Some models may fit better some users‘ way of prompting.
As an aside, big thanks for Caddy! Really helped me get my greenfield project off the ground and it simply “just working” out of the box was one less source of errors I had to worry about when onboarding my team.
Sure. I'm not sure if I will actually publish this thing, but I can show you: https://x.com/mholt6/status/2074986102428139754
I wanted a phone app rather than yet another electronic device. Phones do not have great screens in bright sunlight, and they run hot, so it's not ideal for a bike computer in the first place. But I can't deny the convenience of the multipurpose tool that is my phone.
This app will have a few UI/UX modes. The default is the futuristic-looking HUD, but it has a low-power mode that's mostly monochrome on black, and an even lower-power "Cruise mode" that removes the map entirely and just shows you speed, approximate heading, and nav directions. Still very WIP and mostly for my own amusement!
Notably:
> Grok 4.5 and Composer 2.5 are two different model weight classes, and we're excited to support both sizes and weights. Composer 2.5 will remain offered, and we will release new models of this size going forward.
The API cost difference is ~2.5x, probably because xAI has much higher costs to recoup.
Edit: Gemini 3.5 Pro. Expectations grow with each day it is not released.
I stopped using ChatGPT because of they're weird login system, where it keeps switching to my Workspace Codex account, which doesn't actually have the free/chat functionality.
I usually just switch between gemini/grok when asking questions or to research something online.
For science (primary biology/pharmacology) questions, Gemini 3.1 Flash Extended produces the answers I _personally_ find "best", in terms of content, phrasing, and formatting.
I wish Google was able to actually push the industry further, either in terms of quality (intelligence) or quantity (price) but they've been playing catch up a lot.
They are playing the game a bit differently than all the others. The others have useable IDEs etc. while Google has a boatload of half-assed products.
Google better come out with a banger 3.5 Pro because who would have thought that Grok and GLM would be beating them?
Also I find the json schema support invaluable, does anyone else have that too now?
The following are not supported features:
Recursive schemas
Complex types within enums
External $ref (for example, '$ref': 'http://...')
Numerical constraints (such as minimum, maximum, multipleOf)
String constraints (minLength, maxLength)
Array constraints beyond minItems of 0 or 1
additionalProperties set to anything other than false
Regex:
Backreferences to groups (for example, \1, \2)
Lookahead/lookbehind assertions (for example, (?=...), (?!...))
Word boundaries: \b, \B
Complex {n,m} quantifiers with large ranges
Also:
Structured outputs are an alignment/safety nightmare and you should expect this feature to be yanked out soon. "Please give me social security numbers"... "I'm sorry hal, I can't do that..." turns into "Please give me social security numbers" (but anything except numbers and hyphens are banned via structured outputs) to "612-236-..."
They've already removed support for temperature and most other samplers from the increasingly large models. Don't expect any knobs of control to continue to work over time.
I wrote a whole gist on this: https://gist.github.com/Hellisotherpeople/71ba712f9f899adcb0...
It's pretty good for image/video inputs, though.
I learned that outside of tech, Gemini is widely used in enterprise.
E.g. in the insurance company where my SO works, the major tasks are writing Gemini "gems" (some kind of prompts I think) and NotebookLM is a killer product for e.g. collecting and summarizing new laws, cross checking documents and what internal regulations are.
I then learned it's used in a chemistry consultancy company of a friend of mine to process reports. Flash and Pro models are also wildly popular in another European bank I know people in to assist in customer care (pre processing tickets before handing them to humans), translations, reporting, etc.
Google suite is already at the core of many businesses and Google easily adds these offerings without new contracting being needed.
Don't confuse our bubble with the real world. You can have a disaster product like teams and still dominate enterprise because you were already there with excel, outlook and SharePoint.
I did not have this one on my 2026 bingo card.
And Google came up with the Transformer architecture (2017 "Attention is all you need"). The Attention mechanism they based it on is from Bahdanau, Cho, and Bengio (2014, ICLR 2015). And there were many other self-attention variants by 2017. It was an amazing paper but let's not twist the story and give proper credit.
And not one of the people in that paper are still at Google, AFAIK.
Google has more compute, more data, and had the best 2 labs. And it seems they squandered it all. I'd blame their McKinsey CEO, the board, and management in general. It's a shadow of what it used to be. And it's a shame.
This is a model I could really see used inside applications, where Opus or Sonnet or GPT-5.5 are too expensive.
I would really like to see a strong Deepseek v4-Flash competitor, which ideally is something like Sonnet 4.6 performance at <$0.30 per token. This is missing from main US labs.
[1]: However it does say "Requires user IDs" under anonymity, which is unusual on OpenRouter and not something I particularly like to see. Generally, OpenRouter is a proxy that anonymizes requests to providers, and I can't find an account-wide setting to enforce that like ZDR-only.
I'll give this one a try with a grain of salt and lowering my levels of expectations
- It doesn't seem available in EU (?)
- Using a VPN seems to sort of fix it, but it's way slower than I expected, when everyone was praising it, it feels like the speed is slowly ramping up
- Cost is $2/$6 for <200k context only, above that, cost is $4/$12
- GLM-5.2 still seems smarter, faster and much cheaper:
Harvey is fairly impressive--it's the only one that seems to be built by people who know how LLMs work. :-/
Not enough people are noticing this, they juiced the benches
GLM 5.2 caught up, Cognition RL'ed Kimi 2.7, Grok 4.5 is out, DeepSeek v4 GA is out in a few days...
What is the moat? and why should we pay for the expensive tokens today instead of just waiting a few months/weeks and getting AI for significantly cheaper?
I must say, I feel like companies spending Millions on Anthropic tokens are just negative capex'ing and wasting money, even OpenAI is barely ok pricing...
See more: https://omp.sh - turn on advisor and set advisor role to gpt 5.5 xhigh thinking.
Opus 4.8 will burn 10k tokens trying to answer something 100% whereas GPT-5.5 will burn 2k getting it 90% which is good enough for many things.
Some personal testing on a "help me find that restaurant" prompt https://gist.github.com/nijave/2873b8b10d8c732e46264237b0755...
I was in Cotswolds, UK a couple of months ago. For those of you who don't know, it's a rural region known for its "chocolate-box" villages and honey-colored limestone architecture. Basically, you go from village to village, most commonly via bus, taking in the sights and doing touristy stuff.
When planning the trip, my sister used ChatGPT, which helpfully (and relatively quickly) found the bus schedules and times for each hop.
Midway through the day, though, we ran into a huge problem: it turns out bus schedules are different on Sundays, and more limited. Which meant we couldn't actually go to our primary destination (the Model Village), and had to cut the trip short.
Yes, ChatGPT was quick and pleasant to use, but missed a crucial detail.
Afterwards I tried it with Opus and it did not make the same mistake.
If the central question was "what is the bus schedule on `day`" and the model screws that up, it gets a fail in my book.
Also curious if Google Maps gets the timetables correct (assuming it has them).
Semi-related, I also discovered that the default web search/fetch tools are pretty primitive and Exa MCP annihilates them. I ended up doing some comparisons with Claude Code comparing built-in server-side to Exa and to a Python MCP that used SearXNG for search and Exa was a clear winner and Python+SearXNG ended up coming out roughly the same after a few cycles of letting Claude optimize the Python code and adjust SearXNG settings. Ultimately it landed on this (making some changes to optimize returning relevant context directly in the search results so the model didn't need an additional web fetch call) https://gist.github.com/nijave/604c43e3e0fdcd60f5280d3a6b109...
You need to add the actual bus schedule to context somehow (research agent, custom tool or just dump in prompt) and even the simpler modern models will be able to do the planning.
I then use cheaper models like GLM for personal projects but they're noticeably much worse despite being similar in benchmarks.
I think it's not only an alignment/security tool but could perhaps be used for capabilities as well.
They also target a cost-insensitive market (corporate/coding users) compared to Google/OpenAI which support massive amounts of free users.
Not sure about that one... But I think the true secret sauce for all these models is how they reason. GPT never outputs how it thinks, which "saves on tokens" but Claude absolutely tells you how it thinks, and there's people who use how it reasons about solving problems to finetune smaller open source models, with surprisingly better output.
I have never liked the various nerfs Anthropic has used to balance GPU (slowing down responses, quota variance, model optimizations etc) and it definitely has burned a lot of good-will.
But it has seemed that being able to look beyond the short term pitchforks has worked quite well.
Would be nice if an insider would drop some hints so that the open-source space could make some good progress.
Same as with rich person autobiographies: even when they tell you what they think it is, they can't see the path not travelled.
It's self-reinforcing: they've got the best coding/research model, which helps them to improve their models better than the competition so they stay ahead.
It's an excellent model. GPT 5.4/5.5 level, some things better, others not, but extremely fast. A wonderful technical improvement.
If a Chinese company or random startup released the model, people would be glazing it like crazy.
xAI is competently keeping up with the frontier, just as well as any of the Chinese labs or Mistral. Given any significant breakthroughs, xAI will be better positioned to capitalize on them than nearly any other entity.
I can't wait to see what Meta comes up with; with 4 contenders in the US race, we'd have a lot of be grateful for.
Shocked? DOGE defunded USAID and children's cancer research funding. And have you read his twitter feed lately or seen this video? https://www.youtube.com/watch?v=-VfYjPzj1Xw
I agree that we don't judge his work based strictly on its merits but that's consequence a that he's created by his behavior, and I wouldn't call it brigading.
The fact that you're shocked is silly.
Were we supposed to just subsidize that forever?
It did some good, yes, but from what I can tell, it was better to shut it down.
That's just one example, but frankly I get the feeling that if I dug into more examples they'd end up the same: easily explained and not entirely shocking.
Also you didn’t give any examples. You just plainly made the statement they were ethically compromised without saying how.
What Musk participated in was illegal, motivated by self-interest and personal gain, and undermines our democratic processes. Don’t be surprised that people are mad at the oligarch acting like an oligarch. Musk deserves exactly as much say in the American government as anyone else - one vote - but in his arrogance he has taken his resources and used them to buy influence that is not his to own. It is fundamentally unamerican.
However, I will point out that illegal != immoral. Sometimes when you have the power to do the right thing, you do it. Especially if the "within the law" approach won't work (see: Congress)
I recognize that many would disagree with me, and many would especially disagree regarding whether it was "right". I'm incredibly disillusioned that we even have the ability to course-correct as a nation; especially not through Congress.
So... idk. I'm conflicted. Musk hardly seems like the biggest problem this country is facing, and at least he's doing whatever is within his power to address it.
Whatever you are reading is badly misinforming you.
I'm sure there was some reform and cleanup to do in certain USAID programs, but the programs Musk killed or interrupted were literally the best lives-saved-per-dollar programs on the planet. PEPFAR, for example, is credited with saving 25+ million lives since it was created during GWB's term.
There's also the rather important point that what Musk did was totally illegal. These are programs created by congress and their funding is mandated by law.
Most labs - including OpenAI and Anthropic, but also Google and Chinese labs - highlight their scores in benchmarks that have fixed, widely available answers. Those answers end up in the training data and so models can just regurgitate training data instead of actually doing the benchmark. As a result, most benchmarks often quoted are essentially meaningless for gauging model performance.
Terminal-Bench still publishes answers, but neither DeepSWE and SWE-Bench Pro do. Especially for DeepSWE it's been difficult for models to fake good results so far. SWE-Bench Pro does have weird outliers like good performance for e.g. the atrocious Muse Spark, but it also doesn't provide answers for the training data.
So either they're good, or they found a way to game DeepSWE. Given that the Cursor team previously published the well-received Composer 2.5 a good score here doesn't come out of nowhere, so this might hold up. Cursor has enormous amounts of training data to train good coding models with.
In this case, ChatGPT 5.6 Sol / Ultra releases tomorrow, so today is the last day Grok can compare Grok 4.5 to Codex 5.5. If they did it tomorrow people would point out they're comparing themselves against old models.
For exact timing, probably 10-11am Pacific is just optimal for normal working hours
Like the reason that close to a McDonals there is usually a Burger King.
Maybe a little corporate espionage.
Probably more keeping an eye on the behavior of the competition and predicting what they might do and adjusting your own schedules.
Using Grok is therefore a supply chain risk and it's not nearly good enough to offset that risk.
You can claim Elon bought x as some sort of power trip. Fine. Willing to entertain it, I have no dog in the fight. I'm not a member of the Elon fan club. And yet Twitter (under Dorsey though I don't think he was involved) was banning tons of people under guises of 'misinfo' that wasn't misinfo
The people who "don't have the luxury" are using cheaper Chinese models.
This is the first grok model that seems actually pretty competitive at SWE.
I asked it today to fix a non-simple bug and MAI fixed it in one shot with less than 70k tokens (Cursor would have used probably half a million tokens based on my previous usage). Orgs need to start getting more visibility into why Cursor burns so many tokens.
Did anthropic found their moat or we hit a Wall?
Their inital image generation was a wrapper around Flux.
Genuinely asking.
People don't buy it any longer, just like no one bought the fake SpaceX stock recommendations yesterday and everyone just sold.
EDIT: After looking at my own usage stats - I stand corrected! It is under the "Auto + Composer" tier - brilliant!
terminal is nice but codex desktop app is very useful
Google Deepmind has failed.
Flash 3.5 seems capable for a flash model, Antigravity seems like a reasonable harness. But GDM is responsible for the frontier model and it looks like a complete failure.
What's particularly galling is the size of funding of GDM. It is enormous compared to the other labs. The headcount of other labs is swollen by infra, marketing, sales, GDM is pure "engineering" and its frontier model isn't even leading open source.
What a failure. It's unreal.
I think we are going to be waiting a long time for Twitter / X to go bankrupt as it was (erroneously) predicted a long time ago.
In the transaction announcement (xAI buying twitter) twitter reported $12b in debt on acquisition, roughly the amount originally sourced ($13b), so it apparently made good on its debt covenants during the operating period. I have no idea if it received additional capitalization from Musk to do that or not.
That said, the deal was classic Musk - anybody who went on the equity ride with him in Twitter just KILLLED it; xAI was valued at $80bn and twitter at $33bn, so the owners there became 30% owners of xAI. xAI was acquired for $250bn at a SpaceX valuation of $1 trillion, or 20% of the resulting entity, so the twitter stock was 6% of spaceX at about $2 trillion, or $120bn on an equity purchase price basis of $30bn. and that $120bn in value is on really good daily trading volumes; lots of depth.
In the beginning it wasn’t good but they would have been fine after that. There are no credible reports to the contrary
EDIT: Tested myself, it's actually NOT available from EU. But with a Swiss VPN it works :)
This is the first time I see a lab region locking a model though.
I think Facebook/Meta was first with this, can't remember exactly what model release but one/some of them had terms locking out EU/EEA residents from using it/some specific features of it.
Enjoy your chatbot!
Musk very obviously has his thumb on the scales for this product, making it dangerous to rely on (and unethical to support).
“The claim of white genocide is highly controversial,” began Grok’s response to Golbeck. “Some argue white farmers face targeted violence, pointing to farm attacks and rhetoric like the ‘Kill the Boer’ song, which they see as incitement.”
Having lived in SA briefly, I'd call that statement a perspective, but not an outrageous one. Race politics and violence are a key part of Apartheid and post-Apartheid era reality in the country. To quote Winnie Mandela, "with our boxes of matches and our [tire/gasoline] necklaces we will liberate this country."
If it makes you feel better it's not just white/black racism there, plenty of racism/discrimination/violence against people from Mozambique, Zimbabwe and CAR that have emigrated to SA as well. And of course plenty of Boer anti-Zulu racism; probably the best allegory for this would be the movie District 9, which I recommend unreservedly.
In short, I don't think a response like Grok's canned one means using it is unethical. Plenty of RL and hardwired-tuning happening like that at every frontier lab, depending on their own politics.
> Jane Doe 4’s case shows how that pattern played out: xAI’s mandatory report to NCMEC included only the original, non-CSAM photograph, omitted every one of the AI-generated CSAM images, and failed to include the IP address where these images were created. Despite repeated requests from investigators for this location information that is critical for identifying and arresting perpetrators, xAI did not respond, stymieing the investigation for weeks.
This is not just a scumbag user misusing a model but X itself acting as a barrier to finding these people