A glance at their new index shows that whatever they're measuring, it isn't useful.

Spend an hour with gemini 3.8 and tell me that model belongs in 2026. It feels like the model has Alzheimer's. It gets confused about whether what it reads is what it did. Just crazy bad.

I haven't tried muse spark 1.3. But it must have been a miracle since 1.2 to hit that rank.

Video game journalism vibes all over this.

It's really easy to shit on AI benchmarks, but that noise is useless unless you're offering a solution or a better benchmark.

A projector and speech.

Maybe I'm in the minority here, but I find speech to text / text to speech (but not live audio mode) is quite comfortable and effective for coding now.

The speech to text part can be frustrating if your local tts model does not have word match context for coding. Codex desktop does this remotely well but is slow. I've been experimenting with local software for myself to do this between different llms.

The wall projector is a cool idea because I think it frees the user from staring at a lonely little rectangle while sitting in their fixed office chair.

If done right, this could bring us closer to the dream of more natural, social computing.

Bret Victor's (failed?) project Dynamicland involving a projector on a desk had this goal. I hear he's not much a fan of LLMs. On the one hand, I can see why. But I think, used correctly, it might be the sort of thing that unlocks his dream and, really, my dream, too.

A here's a presentation of Bret's talk on it: https://www.youtube.com/watch?v=7wa3nm0qcfM

Slight tangent: using speech to text to ramble about your rough design for like 20 minutes to an llm produces surprisingly good results over short prompts even when you contradict yourself. They're so good at picking up on what you're orbiting.

What I saw was multiple people living alone in a small box in a warehouse (probably filled with other boxes) with all of their natural, social interactions directed at a wall. I wonder if this is foreshadowing for the future of work, at least it is what work will look like as envisioned by OpenAI.

Of course they're mercilessly mocking you instead, but hey.

It is still a quite fast. In fact, I just typed this in using this app.

Governments are defined as having a monopoly on violence. And we, through a lot of struggles, were able to attenuate this power such that this power was not granted to a single person alone but both distributed to the judgement of many and codified so that, in theory, no one is above the law. Even imperfect, this cultural belief that governments should behave as such has caused it to be, in a flawed way, more true than it used to be.

These digital gate keepers have special powers that control identity, the algorithms deciding your information awareness, and access to what society now requires to live a normal life. So, what are they? And where is our enlightenment stage? What sorts of human rights are there in this world that we agree on, at least culturally.

I don't believe government regulation is the answer - it has both failed and centralizes again in fewer hands. The thing we need is cultural understanding first of this new sort of entity. And an ethics system that we can agree on that can constrain them as the Magna Carta once described.

How should our collective values shape this?

https://en.wikipedia.org/wiki/Magna_Carta

A more realistic approach is legislation that either forbids providers from suspending or otherwise limiting access to any government integrated system or simply forbidding these types of integrations altogether.

> A more realistic approach is legislation that either forbids providers from suspending or otherwise limiting access to any government integrated system or simply forbidding these types of integrations altogether.

There's more history in the US of successful legislation aimed at breaking up big companies than at that sort of specific narrow technically-aware well-executed regulation.

The former is blunt force. It's relatively easy to target and make effective. It just needs enough ground-level support to get enough people from both parties on board - and we're seeing some shifts that way in the current Republican mishmash of groups.

The latter needs excellent technical knowledge and resistance to super-sophisticated lobbying and loophole-creation. It's very hard to get right.

We live in a much different world today than during the trust busting era. Extreme concentration is the status quo of today's economy and it's happening across basically every sector and industry.

Lina Khan as the head of the FTC under the most legislatively prolific Democratic president of the last 30 years was completely powerless to break up anything. Even Trump and an army of partisan DAs that were frothing at the mouth to dismantle Trump's enemies have failed to do so.

Just yesterday Google prevailed against the threat of a breakup of their ad businesses which is perhaps the single most compelling example of anticompetitive abuse by big tech besides maybe the Apple ecosystem.

This trend is only going to accelerate as skilled workers age out of the economy and white collar economic activity aggregates into a handful of massive tech companies.

That's all without factoring in that government quid pro quo is now defacto legal and these companies have very deep pockets for donations.

There are many other factors I could mention but the point is that nobody with any power wants to break up these companies.

Only because she was blocked by both Republicans and Democrats who are sold out to the same plutocrats. I agree with the parent that the problem remains very simple and is a matter of balance of powers. Any fine-grained regulation will be simply ignored by those people who are above all laws.

EDIT: Maybe i was a little too lenient to her. I mean, if a politician really wanted to change the status quo, they'd appeal to the public to help change the balance of power, which Lina Khan never did.

But who is the 'we' in that 'we'? If you defer to a politician, you simply centralize more power into a body of government constrained by only cultural norms and donor interests. Making a change requires the shifting of the overtone window of acceptability. At that point, it executes on it own regardless of mechanism.

The cultural norms are the prime mover.

You're taking it as a premise that someone should have this power, i.e. that the centralization should actually exist.

Compare two alternatives.

One, each person has (is required to have) a single account across all services. You use the same account for the DMV, YouTube, your bank and the IRS.

Two, you have an independent account for each service. You have separate accounts with the IRS, Medicare, DMV, YouTube, etc. If one of them gets messed up or compromised it doesn't affect the others. If you have a dispute with any of them it limits the blast radius.

The first one is a dystopia regardless of who controls the single account. It's subject to abuse of power whether it's a politician or Mark Zuckerberg. So the solution is to demand the second one.

But you're still relying on a law if I understand you correctly.

If the society lacks them as deeply ingrained values as opposed to somehow to getting lawmakers to force them, then they will be implemented cynically and not within the spirit of the values. The type of thing that happened with apple's implementation of the anti-gate-keeping laws 'enforced' by the eu.

What you describe requires a flexible values system - not a single semantic standard. I'd love to live in a world where if you were big enough to have people wanting to interoperate with your service on behalf of your users and you were actively preventing it, you'd be an outcast in the deepest sense. That no users would touch your service, and it would be politically unfeasible to deal with you.

We don't live in that world because people have not developed this sort of social norm.

We currently have an army defeatists who show up to tell you that if you think companies would get more customers by selling products that screw them less, nobody cares about that and people just want to live in a convenient cage with an unlimited supply of artificially sweetened penitentiary gruel. These people (to the extent that any can be found who are not robots) should obviously be burned alive in the town square to serve as an example to the others, but the most important thing if you intend to solve the problem in that way is to ignore them and build the product that screws customers less anyway. And then shame the services not doing that into irrelevance, once the alternative actually exists. Put it in your advertising so people know there is a difference.

Likewise, there are better ways to do anti-gatekeeping rules than what the EU is doing. For example, require two things. First, that source code be published as a requirement to receive copyright protection -- not that the customer can make unlimited copies, but that they can read it to know what it's doing or make changes to their own copy. And second, require that if the manufacturer/seller can modify any code running on the device, so can the customer. Now Apple can do malicious compliance all they like, but any vendor locks they put in will have someone distributing a patch to remove them the next day, and inhibiting that is the thing which is prohibited.

I assume that by this you mean something more nuanced than "we should abolish the FAA and let the airplanes crash until private enterprise solves it." Or maybe "Let drug dealers (pharmaceutical companies) sell whatever the free market will bear without testing, safety, or the ability to hold them responsible."

I only ask, because there are some people who unironically think that we'd be more safe going back to the time before OSHA, or medical licensing, and some of them hang out here.

Strawman fallacy: It is in the interests of private enterprise to not have their $100-million+ airplanes crash, because it means $100-million+ going up in smoke. The owners of those planes would therefore have an interest in not having them crash.

> Or maybe "Let drug dealers (pharmaceutical companies) sell whatever the free market will bear without testing, safety, or the ability to hold them responsible."

Under this scenario, if the public finds out that a drug is harmful/lethal, the public jerks away from that drug towards solutions that have a reputation of working.

Is the resulting rate of fatality/worse-QOL lower than (a central authority filtering out drugs & holding them to standards)? No, definitely not. But the gap between the two is not infinite, and in most cases falls into an order of magnitude of difference.

-----

> I only ask, because there are some people who unironically think that we'd be more safe going back to the time before OSHA, or medical licensing, and some of them hang out here.

An axiom of (feeble-mindedness in the average person) is proposed here. People are as good/competent as the standards that they are held to.

> medical licensing

There's a difference between a certification of competency, and a deliberate constriction of the supply of medical residencies.

https://thehill.com/opinion/healthcare/5550556-residency-cap...

Rich people and fascists don't want a smaller government, they want a government too weak to protect you from them.

You seem to be saying that there are a group of people who want it this way and they have more power so that is why is it the case.

But how did we get here? It's a complicated question for sure. And for anything that occurred like this, there are many answers depending on where you put the weights.

I put the weights cultural norms of the maybe 100s of thousands of people that had a hand in how we got here in varying degrees. The leaders of the companies of course who didn't get flack in their communities for doing something that was negative to the rest of society. The politicians who played along, the employees who were proud to work at such a place and the family that could be proud of them for it, even us users who sacrificed liberty for safety and convenience.

At every level, it's social norms that caused this to be.

We are, primarily, social creatures. Even the rich - maybe especially the rich - are greatly influenced in how they act by what those they respect think of it.

It's still about framing, though. It's not a fight. The gatekeepers, like governments, laws, and money are social constructs - shared fictions. They are not physical things but made agreements by society. We can't fight them anymore than we can fight ourselves. And they stop working the moment we stop believing in them.

The only way to fight a social construct is to create a new shared fiction that is more desirable and works better.

The machine created by the current social fiction is composed of people who would all be better off with a better narrative. We just don't have the imagination right now to see it. But it is there waiting for us somewhere. Like all the best ideas that transformed us for the better throughout history.

While I agree with where you're coming from in spirit, if you want to push this then you've got to do the work of actually coming up with workable concrete approaches rather than hand waving against a whole general category of action. The problem is that digital communications lend themselves to centralized power, and power vacuums inevitably get filed by something, right?

I too would love to see the power of surveillance companies get magically reduced by something that isn't centralized government action. But that hasn't happened so far, right? If anything, they've grown stronger - sucked in ever more of our lives, sucked in technical communities who were culturally predisposed to and capable of resisting, and even taken over control of parts of the de jure government!

So barring that, government regulation (ie democratic accountability) seems like a necessary approach. This fails too of course (eg how all the natural rights referenced in the Bill of Rights are being infringed in one way or another). But once again what is the alternative to attempting to codify lofty ideals, writing them down, and then pushing to implement them wherever they're being infringed? Regulation like the GDPR and other EU directives at least attempt to draft rights-based paradigms, as imperfect as their implementation and enforcement may be.

The real problem would seem to be that we're soured on the idea of having reasonable laws in the US from every attempt at meaningful regulation ending up with a multitude of holes in it due to corporate lobbying. It seems like that is what needs to be focused on and addressed - thinking one can let the corpos run wild and then somehow outrun or individualistically push back on them is a pipe dream.

1. I think sometimes the shape of a technology pushes a shape of society. But there's a feedback loop here. The like 15 years of tech were dominated by ad-tech which lead to algorithmic feeds that made engagement the game. This had an accidental effect of enraging people against eachother. This all started because of how much money google started raking in money on ads - people were aware, and we copy. Tech's focus on ads was an accident of history which could have gone other ways.

Maybe this seems intuitively wrong. Like there is some evolution about how these things must go. But there are a lot of examples in my experience that people see something successful and copy it and then it turns into a cottage industry of whatever that community was.

The best way forward here is to build and be successful while still following good values. The tech moment right now is ripe for it. There's a lot of reasons why open source, local first, and catering to the end user (and not making them the product) is a great way to be successful starting a company right now. Being successful in this way and talking about it breeds more likewise.

2. From what I can tell, pessimism is actively harmful. I've spent a lot of time with Russian communities (most who left due to the war) in these past years. That got me very interested in Russian history.

The weird thing about the arc of it is how similar it has been in every government they've transformed into. The same basic values hold which leads to the same problems: centralized authority, low value for human life, and general oppression. This is true in czarist times, or soviet times, or post-soviet times. Government change doesn't change cultural norms.

That's not to say I think that they are forever doomed. But it is the general zeitgeist of society that leads to the same place. And for it to ever change, their cultural values need to change.

There's a parallel universe out there where Russia is the richest country in the world that enriches and benefits the rest of world as a partner. This is obvious given its natural resources and geographic location at the nexus between great markets, and its history of producing terrific scientists and engineers.

It is the pessimism that kills all of that ultimately leading to autocrats that live to hold on to power at the very high cost of the rest of the country. The assumption that the world will just continue to be shit is what causes this cycle of abuse. It's a very hard cycle to break out collectively once entrenched. As all collective action problems are. But my, now longwinded point, is that it is not government that is the mover of change but philosophical attitude.

Zero-sum thinking, tribalism, and low appreciation for individual self-determination turns us all into a raging bucket of crabs. There will be a king crab but his life will suck too.

I completely agree about the feedback loops. They're mutual feedback loops though, and causality can go both ways. We're currently stuck with a strong attractor of surveillance-based software. It's certainly possible to escape from the local minimum, but requires a lot of activation energy to escape it (compelling functionality, features, ease of use, etc, to prime its own network effects).

I don't think it was ads per se, but rather the centralizing software architecture of AJAX (et al). Everyone wanted to build things that were "easy" for "everyone" to use, how to make money came later. Or going back even earlier, if "the web" had been designed around something like content centric networking rather than HTTP.

On "tech moment right now is ripe for it" - for me, it has felt this way for the past twenty five years! I generally agree about pessimism as well, so let's just say at this point I'm tired. I certainly don't want to dissuade you from pouring new energy into software that is going to change the world [for the better]. But I'm also cautioning you to not reject other angles of reform!

Further on the pessimism point, isn't shying away from government (ie collective) action itself a form of pessimism? I'm pretty pessimistic about groups (herds) of people in general (things always end up going "sideways" due to opportunists). But looking at the structure of US politics it seems there is a large contingent of "government can never work" that ends up running interference whenever de jure government tries to reign in the corpos (which tend to act as de facto government), creating a self-fulfilling prophecy.

Or putting my general point another way, in terms of the culture point you made elsewhere - our mass culture is currently strongly steeped in centralized political action/narratives. If you want our mass culture to change to value technical decentralization and privacy, then one of the main things you'd expect to manifest itself is a call for centralized/government action, ala privacy regulation, right? We can certainly lament the centralized political culture, but unless you're talking about changing two aspects of our culture simultaneously, then that basically is what it is.

I don't think shying away from governing hierarchies are pessimistic but maybe you see it that way because I didn't explain myself well.

It's important to separate something. Hierarchy and cooperation are not the same thing. Though I guess in instantiation at the human level, they seem to mostly be.

But multi-cellular creatures are made up of cooperating cells wherein there is no centralized power. It seems like nature has borne out that this configuration is robust.

And, if you squint, markets are a similar thing. I love markets. They're a way that people can make individual choices about what they can offer their neighbor and so be rewarded by their neighbor in this positive sum way. It is, by nature, decentralized (though regulation exists in modern markets and sometimes work well for a time as a patchwork kludge over a janky system).

Anyway, there's trust networks there and lots of interesting stuff we could tangent about. But the point is that markets are a decentralized thing have been around longer than governments. We don't need monopolies on violence to cooperate if social norms forbid violence. Of course, the violent ones hacked the system and so we needed to defend against that, so we picked the violent winner, and governance is born.

This is also an interesting path turning point that, given our current technology and understanding, if we started from zero, you could wonder if we could develop in a way without a single entity with a monopoly on violence because we handle this sort of threat somehow else.

There's that whole debt and reciprocity thing that is even older than markets - hell, animals do it. But I wonder about the scalability of that. I don't believe david graeber's idea that the abstraction of reciprocity was the root of all evil but that's like one take and maybe we could make better reciprocity networks.

This is all rambly. I hope my point comes across though. We lack imagination for configurations that can be that are cooperative without the cancer of centralization.

It would be nice if someone starts imagining this for us. It's very hard to imagine a world so different from our own. But we know, by nature, that it is possible. There are a lot of cells in my body - it's not a scale issue.

My take is that cultural values/social norms are the real thing that makes society. We think it's all these things like laws, money, and borders. But those things are secondary fiction memes to the primary fiction meme of values.

The thing about values are that they are very resistant to centralized control by nature.

1. Assignment of addresses is centralized. Even though it's federated (ie distributed), authority still flows logically from one single authoritative view. And even though individual networks can choose to go against that central view, diverging too much prevents them from talking to one another (ie most of the bargaining is on a take-it-or-leave-it basis, rather than something more nimble or adaptive)

2. It's a protocol for communication between nodes. Reusing node addresses as service/contents addresses leads to services being inherently centralized on those (logical) nodes. Not a problem with IP itself, but definitely an attractive nuisance that made problems for things naively layered on top of IP like HTTP(S) and DNS.

Additionally, they are unelected.

Whenever I think about AI taking over the world I think about this really neat 1970 movie 'Colossus: The Forbin Project', https://www.youtube.com/watch?v=qoB1l3A-GF0

I call it pissing in the pool.

I hope one day we see paid advertising as immoral. Immoral acts can be defined as negative sum on society but possibly positive to the individual. A working society agrees, culturally, to think negatively on negative sum behavior.

Social animals do this too - even vampire bats ( https://pmc.ncbi.nlm.nih.gov/articles/PMC4685829/ )

The issue with paid advertising is that it is more difficult to understand how it is a negative externality. There's also the fact that so many make their living on it these days that it's tough to get them to believe they're doing something wrong.

But imagine all the distraction and labor we could avoid if we could agree that it is not right.

Advertising is the always hungry beast that even a post scarcity society would starve feeding.

I did contract work so I could stay traveling and lived on money I made in previous full time jobs. It was a very nice way to live compared to the 9 to 5 even though I made less money.

Having enough money to live the way you want is all you need. Hostels are fun.

But in the free time I had, I did end up making a SaaS that set me up very comfortably for the rest of my life.

Cynicism will keep you in your 9 to 5.

Hostels are not fun, unless you are in your early 20s.

If you don't, even 5 star hotels can be miserable.

I find quieter hostels much better. When I was over in Ireland, the ones in Belfast and Dublin were awful, and in county Galway. I found wonderful ones in Co. Donegal, Armagh and Limerick.

I've had noise issues in hotels and I've had quiet hostels. Totally orthogonal.

That said, much like hotels, some hostels do suck; bedbugs are not chill, location is important, cost is important, vibe is important. I found that hostels that were more suited to families trying to save money were the least appealing as someone with just a spouse and a type B personality.

I hate paying for hotels so much that I visit my hometown less than I would otherwise if the city had a decent hostel.

There were examples like this of people in the groups hitting on something and making enough money to retire (if they managed their money right)

It was the rare case. The end result for most of them was moving back home and taking a normal job when hopping between hostels and digital nomad group homes stopped being fun and none of their business ideas got traction.

Models get confused by who said what - especially cluade models. They get confused by negation (don't do something versus do something). Compartmentalization is hard.

You can either solve compartmentalization completely, or just not tell the model to do things that must be compartmentalized at high stakes.

One great thing about the rise of AI is that now it's super easy to make and maintain integrations. This is a place where AI coding uniquely shines due to the points being well defined and mostly quite standard.

This is going to hurt the monopoly power of software companies. But it'll be great for allowing the proliferation of smaller companies, open source, and home grown solutions.

I'm already seeing the pendulum swinging back to local first instead of cloud services.

A lot of SaaS companies change their API in subtle ways throughout the months probably too rapidly. GraphQL schema changes, splitting apart one api endpoint into 2, changes to authentication method required. A thousand little cuts.

This used to be my biggest time waster. Now it's barely on my radar. Life changing for me. I guess your mileage may vary.

Concurrently, the writing is on the wall for us SaaS product providers - lower margins, and more competition for the same reasons. I welcome it. More quality and choice at a better price for everyone.

Metrics can only capture short term value.

They also let us make unethical decisions with the data taking the blame.

It is fundamentally dehumanizing. We are no longer treating others the way that we would like to be treated when we make software.

AI is a printing press or a steam engine. It's not a weapon.

The framing is important. Much like 'distillation attack'. These framings are an attack on the narrative.

Printing presses at the time were seen as extremely dangerous too. For many of the same reasons cited about AI.

In both cases, the technologies were used to destablize existing power structures. And, after some hardship, make a better world for most people.

Even calling these things 'dual use' is a weird framing. Books are not 'dual use technologies' unless you want to sell fear.

I've not heard of a printing press going rogue and attacking things.