I hope the future of AI isn't this sort, where companies provide the user/customer with an interface to an AI that can do things for the user... I would much prefer that companies instead provide an interface FOR an AI, and the user brings their own AI which connects to that interface.
In other words, provide my AI with tools, instead of providing me an AI that uses your tools.
That way, my AI can bring all the context it needs, and I can bring all of the settings and knowledge about what I want with me. I don't want a fractured world of tons of AIs i interact with where I have to explain all the fundamental information about what I want and how I work every time.
This also has the benefit of sidestepping the issue the essay is talking about. You provide a consistent tool, and the AI weirdness is not your issue anymore. You don't have to worry about solving for all the weird ways people prompt the AI, or the ways they break.
There's no interop between video conferences services, no BYO-video-conferencing app, why would this be any different? It took government regulation to allow telephones other than those provided by the phone company to be plugged into the phone network. Can't control you or usage if they just provide dumb pipes/interfaces that anyone's tool and interact with.
All of those browser and computer use agents (available from OpenAI, Anthropic, Meta, and X.ai already) represent an interoperability workaround already.
I think this is basically the motivation behind MCP servers, so you're in good company.
But I've spoken with many people at companies who've decided to add an in-UI agent to their apps, and I don't think this trend will persist. Absolutely AI/agents will increasingly become a major component of interfaces, but the current common incarnation of "Clippy for X" feels like a kludged bridge between an app that was not designed from first principles for agents (and often was poorly designed for humans) and the impulse to be "AI native".
From my experience, some of the most successful niches for this sort of UI so far have been in apps that were already well suited to it. I'm thinking specifically about analytics/dashboarding/"this is a portal for you to query things easily" software. They already started with a lot of the elements you want: visible provenance of the agents actions via the queries it writes, a malleable interface that you already expect to be customizable and ephemeral, and most importantly, "navigation" that is genuinely difficult for many users (in the sense that "navigating" can mean "querying specific data"). The agent provides a ton of value to users and its actions are intuitive and legible.
But the agents you see right now in a lot of apps that do things like navigate you to the right page by... sending you a link, which you could have clicked from the navbar? Or worse, which hijack your navigation and throw you on a page you're unfamiliar with, and where you have no sense of place or how to make your way to/from? I don't think they're particularly long for this world.
IMO it's a fundamental tension. AI does not fit well inside a product. It works best outside of it. But that means the product - not its functionality, but the business around it - becomes irrelevant. Which none of the product vendors want.
Also the motivation behind CLIs and SDKs, right? And the way agents write Python regularly, these days, I am hopeful what can be offloaded to deterministic systems will be.
A better internet is one in which LLMs enable everyone to ride their own custom Chromium (or whatever) flavor and the "frontend" of > 90% of sites - especially e-commerce - is generated specific to the user given their configuration of their personal LLM intermediary.
Companies will increasingly not ship frontends anymore, maybe just themed components that most users will ignore. "Pages" naturally become synthetic amalgamations of the various things you are interested in from various sources. Your preferred UI/UX comes for free courtesy of your LLM intermediary.
Everything is an API. Regulation will probably be needed to enforce this on the data access side given the incentives for companies to resist despite the benefit to the consumer. A relatively easy solution is to make it part of the requirements to take payments online, which is already highly regulated. The goal would be that essentially to operate anything approaching e-commerce online you must allow consumer AI intermediaries sufficient access that they could independently construct whatever your current frontend is from the publicly available API/MCP endpoints.
The tricky part of engineering around privacy/personas, account creation, etc. will be relatively easy following such state intervention. Take payments - virtual credit cards are already a useful way to protect data and mitigate risks when dealing with multiple vendors. This sort of API is a natural fit for a world in which companies are not allowed to box you into their horrible UIs, most of which are simply lazy attempts at copying the most profitable anti patterns of their competitors. Let your personal agent generate a unique card for each vendor behind the scenes. Why should you have to care?
How much better the internet would be if every UI/UX was in a meaningful sense, your own.
I feel like this is a bit too-optimistically describing an outcome which is unlikely, given how consumers have been slowly losing the same war for decades already.
The elephant in the room and the really "tricky part" involves powerful advertising and copyright-holding companies. They will (continue to) interpret that kind of user-agency as an existential threat. Their businesses revolve around deciding and knowing what you actually see and do with your computer, and their toolkit isn't just technical, but also economic and political.
That's why we already have problems with DRM, "felony contempt of business model" under the DMCA, browser fingerprinting, and everybody surrendering to arbitrary JS code controlling their computer. We capitulate not just to see an article or talk to a friend, but also to avoid our technical details getting into a secret un-appeal-able "suspicious" blacklist.
That basically describes the Google shopping ribbon that appears with searches for products. When Google was good, you'd say a website doesn't exist if you can't find it in the 1st page of Google. The shopping ribbon provides offers from websites that aren't on the first page of my search result. Although I don't think you can pay without visiting the seller's website, this means that these are effectively stores that are more visible through the APIs they use to talk to Google's recommender engine. These suggestions are also no doubt personalized.
This also describes adsense, or Amazon or AliExpress, so it is clearly a successful model. If there is much left of other models. The only thing we have to hopeful about is that the interface is standardized so commodified SWE can be used to make your own Amazon or AliExpress.
You shouldn't need LLM and AI to do such things. Even if such things might be useful to some people, I might want to use the API without LLM/AI and just to read the API documentation and then to make my own, and some other people might want to do similarly. In the specific case of e-commerce, I had been making my own specification of a file format for e-commerce (which is independent of the protocol), which is not intended to use with LLM/AI, but instead is intended to avoid many kind of dishonest business while also being flexible and that you could make and use your own software or some other implementaiton instead of being forced to use their UI/UX.
We could have a sort of “Representational State Transfer” using HTTP to represent the state of resources in both a human readable and machine readable format.
Some companies are taking the opportunity to use the MCP server to send messages to the agent that the user won't necessarily see to protest. Agents are generally incredibly gullible and a huge opportunity for this type of thing.
I'm currently building something like this, basically converting every service to a standardized API that you can consume in a structured manner. Not focused on AI, but focused on normal users with special requirements on usability.
Basically the ad blocking approach has failed, and one needs to recognize that an allow-list based approach is needed when consuming internet-based offerings. Pick the cherries out of each website, leave the trash behind, and then build a custom per-website GUI around the things that you actually want to consume.
I don’t think that will happen. It sounds good, and I would like it if things operated that way - but from the company perspective how do they, for example, have a tool call that gives the customer a discount, without it getting used when it shouldn’t?
Companies want ai to replace human customer service decision making, which means it can’t just be an api that an external agent can interact with, because it needs private knowledge of company processes and access to capabilities that are abusable.
But we’re already at the point where if you manage to talk to a human, mostly you end up speaking to someone with no actual power to resolve your issue - so i think basically the future is just going to suck
> from the company perspective how do they, for example, have a tool call that gives the customer a discount, without it getting used when it shouldn’t?
I might be misunderstanding your point but I think you're in agreement with the gp, as in they are arguing for a locked down interface behind which the AI sits, narrowed to use within whatever particular operations the user can access, as opposed to a wide open AI interface with access to any operation (in principle).
So in your example there's no way to ask the AI for any old discount because the interface for doing so is locked down to just the discounts you can access, although the AI may be able to apply a discount you didn't ask for dynamically, if allowed, to give the customer a better experience.
Of course both versions can be implemented securely, it's just probably in general a smaller attack surface if you provide a tighter interface first.
> how do they, for example, have a tool call that gives the customer a discount, without it getting used when it shouldn’t?
One possibility is that the company and the user both have AI agents, and that these are able to negotiate with each other. Then I talk to my local customized agent, which knows my preferences, and it explains the required context if the company’s agent does something that won’t make sense to me. But the company’s agent is still privy to the details required to offer discounts, say.
> have a tool call that gives the customer a discount, without it getting used when it shouldn’t?
without comment on the rest of your post, a lot of discounts are just dark patterns, buy 2 for $10 or one for $6 is one of the oldest dark patterns around, and a lot of people fall for it, when they really didn't need that much whatever it was they were buying. In fact there are laws outlawing these practices coming into force for various types of products (alcohol, sugar) .
Other types of discounts like loyalty are more vague personally I think are a minor dark pattern. Negotiated discount on bulk purchases b2b are a different class, but it could be argued that you should only get the discount that the seller saves on logistics.
I'm just pointing this out because its very normalized, 'signup and get 3 months free', 'buy 2 get one free' etc, people dont bat an eyelid at, the same as advertising , but they are all just the earliest versions of dopamine hacking and dark patterns that consensus now is coming around to say probably isn't the best thing when scaled beyond a simple one to one interaction. Why are snapchat dopamine hacking with streaks but your local coffee shop stamping your card for your 10th coffee free different?
> provide the user/customer with an interface to an AI that can do things for the user
This is exactly what MCP is. But the reality is it will likely be about as popular as browser extensions and most normies will avoid.
Simplifying UX at the cost of personal control is inevitable because not everyone wants to think about tool selection and coordination. But maybe we can angle the future toward "tool bundles" that interoperate well or (if we are dreaming) mandate models remain accessible by any harness, which none of the players want but would be best for users and ecosystem development.
> where companies provide the user/customer with an interface to an AI that can do things for the user
I'm afraid it is the future. The companies will protect all documentation and IP by putting it behind AI agents, and will monitor (with another AIs) how is it being used.
It's gonna be a dark era for any knowledge in public domain.
At that point, companies could just provide an API and let users bring their own user agent, be it AI or traditional.
The problem is that history has shown that companies really do not want to do that. They will happily pay to maintain an inferior interface and go out of their way to disrupt third party clients.
If the interface is documented and can also be used without AI, and you can write your own software instead if wanted, then that will be more helpful, I think. (Someone who does want to use it with the AI can still do so, without forcing everyone else to also do.)
This is like asking LG to make TVs where you can bring your own video input. It just doesn't make economic sense when they can extract so much money by owning the glass.
100% agree, and I predict this is where the actual nearest big battle in our industry will occur.
AI subsumes products. Users want that. Vendors, do not. AI does not work well when shackled within confines of a product - it works better from outside, where it can treat slices of products as tools, and mash them together into ad-hoc solutions. Alas, products is how our industry makes money. Take arbitrary slice of problem space, slap a trademark on it, and shill to people (or VCs for funding). This disconnect makes AI an existential threat to a good chunk of software industry, and you can bet that companies (possibly including your own employers) won't go gently into the night.
MCPs are an aberration, the early stage of "AI adoption" where no one knew what they're doing but they knew they "have to do something with AI!". This age is now ending, and I expect the tensions will go high, as most vendors will get desperate to avoid their products getting obsoleted overnight.
Imagine if companies exposed access to their services via a discoverable uniform interface.
Each company would determine the range of actions a customer could perform in response to their request and current circumstances. A representation of state could be transferred between the company's server and whatever client the customer chooses to use. The customer would be free to interact with any of these services in a way that suited them via their chosen client.
Either that or we could kludge something together with MCP.
So I actually struggled with this recently (2 months ago). My "vision" was to have customers BYO agent and provide a secure MCP with an in-app (forced) approval gate for sensitive operations. But everytime I described to one of my clients, how they could bring their own ChatGPT, Claude, Gemini, w/e, I got blank stares.
So now I'm building (already have done so mostly) cheap ZDR models into the application. I'm trying to build them in a way that I would want to use them, so a less "I'll be your AI today!" help bubble, but who's kidding... At least they aren't just customer pacifiers that use the help system, but actually can act on behalf of the user.
After which I'll end up building the MCP, but for power users, hopefully leveraging much of the same work. It's worth noting that MCPs have way too much friction still. It's gotten a lot easier recently (very recently) to bring your own custom MCP with Claude, but ChatGPT is more work, and restrictive (specific account types or submitted MCP apps) for the end user and I don't even know what Google wants at this point... It's a bit messy. IMO they are all dropping the ball (except Claude, which has taken the best approach).
I think we might be in this interim state that requires us to build ALL of the UIs, to address ALL of the users (Human responsively and Clankers), and it's a bit painful. If I was "the user" myself, then I would just want a CLI w/ OAUTH with clear documentation, that an agent could use (how is this worse than MCP?). But I'm not my own customer. That being said... I probably WILL build this version to scratch my own itch and bet on the future.
You don’t understand fondly reminiscing about when the world seemed more magical? It has nothing to do with the truth of it, it has everything to do with how it felt.
This reminds me of when I coach a kid in chess and the first thing I ask them is what they want to achieve. A fairly common response is to not lose. At first that doesn't sound unreasonable, yet the goal there doesn't require any coaching at all - simply only play terrible players!
So too with knowledge. The path to never believing anything false would be to simply minimize the amount you learn, and certainly to avoid learning anything developed within the past century or so. It's not a great perspective for self betterment. Learning from one's mistakes is the path to improvement in most of every domain. Mistakes and falsehoods are nothing to regret. If anything, they're something to rejoice over. It's only the fool that believes all he thinks must be true.
Hmm. I tend to have the opposite. With science usually the process of overwriting a credence means increasing understanding and getting a deeper perspective. No one is born knowing anything, and as we advance it’s inevitable to learn certain prior beliefs were wrong. Might as well make it a fun thing.
The most polished Glacier backup/restore mechanism I know -- because I helped build it -- is through rustic <https://rustic.cli.rs/docs/commands/init/cold_storage.html>. It has native support for backing up to Glacier (because writing to Glacier is identical to writing to regular S3), and it supports external "warmup" programs to support cold storage of your choice, with Glacier having an existing warmup program you can use.
You can exercise it against regular, non-Glacier S3. Same flow; the only difference is that you skip one attribute on upload and one request/wait step on download.
How do we determine at what point something is “too addictive” to be allowed?
There are many people who can gamble responsibly and enjoy it. I love playing poker, and play in a tournament every few years. Lots of people are compulsive gamblers, though, and lose it all.
Do I have to give up my hobby because some people can’t handle it?
What about drinking? I enjoy a happy hour once a month or so, and have a couple of drinks when I do. Others drink 10+ drinks a month. Do we ban it because of the people who can’t stop? (We know how well that works)
I am not saying we do nothing, but I don’t think it is as easy as “just pass legislation”
We know how full 100% alcohol prohibition in 1920's worked in the USA, but what's missing is that it doesn't have to be a full 100% for it to benefit society. Laws restricting sales, like no selling alcohol past a certain hour, or on Sundays, still exist on the books in many places, and while I can't speak to how well they actually work, given that we'd just stock up on Saturdays anyway, there's also no Al Capone levels of violence in places with those restrictions. So let's not say "we know how that goes" hinting at 1920's prohibition, which was over a century ago now, and just leave it at that.
For a counter example, I like driving fast. Really fast. Should I be banned from doing that because other people can't handle that responsibly? Society has looked at that and decided that, yes, no one should be allowed to do that on public streets. So I keep that behavior to the race track and sign an extra liability waiver in order to do so.
Why should gambling or alcohol be treated different from that?
> Society has looked at that and decided that, yes, no one should be allowed to do that on public streets. So I keep that behavior to the race track and sign an extra liability waiver in order to do so.
So can I still drink and gamble in private if I sign a waiver?
What would a gambling liability waiver cover? Your losses/damages are not in a confined circuit or ringfenced from your general finances. Are you just agreeing not to sue if you lose your house/wife/job?
Sorry, but alcohol prohibition did not work and was instrumental in fueling the rise of organized crime (does that ring any bells?).
I happen to believe the War on Drugs (prohibition v2) is an even bigger mistake than the first one, which also ignores why it came to be: v1 prohibition enforcers needed something to do and there was a growing problems with an "uppity" segment of society that needed to be put in its place. The history of it is horrifying; and the current implementation of it is horrifying as well.
So legalize it all. Regulate the advertisement/promotion/access/purity, etc. It's going to happen one way or another and pushing it into the shadows only makes it worse.
So let the gambling happen but keep it on a short leash.
If all you're saying is that it needs the right legislation/thoughtful legislation and not just a blanket ban, then I guess I agree. The base of this thread was someone saying that all gambling should be state owned, which seems like a reasonable thing to me.
On the other hand, I am also looking for a contractor to do my concrete. I hate the whole process, I hate calling people only to spend 5 minutes trying to explain what I want and then they don’t provide that service or try to convince me to do something else, and then seeing bids and not really understanding what things are important and if I am getting a good price.
I would love for an agent to just be able to handle all this for me. Find me a good, fair price from a reliable contractor who understands what I want. I don’t want to have to call a bunch of people, just do it for me.
I tried seeing if an agent could help me with this, but it isn’t quite there yet. It did a good job helping me figure out what I really wanted, but not the rest of it.
I got quotes first which was the wrong approach. I ended up taking photos and detailing what I want on plat of survey. I also consulted a structural engineer(for garage cracks) and researched using rebar, fiber mesh, and metal mesh.
So, basically I know what I want and it is easier to get quotes this way. I just send the docs and dimensions.
Well, now I have the same problem I had before, how do I find a good structural engineer, and what is a fair price to pay them? Do I have to call a bunch and do the same thing I hated doing for the contractors?
I would call 4 local structural engineers to you. Ask about price. Inspection and report is about $700-$2000. You just go with someone who feels right for you. Still cheaper than $40k concrete job, so not as much of a deal.
The red scare of the 1950s was ridiculous enough, but at least the Soviet Union was an actual threat. We are now at a point where “pro-transgender” is the new threat?
Sure, if you conveniently drop the qualifying word "radically". But that's not what was said. What was said specifically included the word "radically" for a good reason. Radical anything is a threat.
> Poe's law is an adage of Internet culture which says that, without a clear indicator of the author's intent, any parodic or sarcastic expression of extreme views can be mistaken by some readers for a sincere expression of those views.
The administration doesn't like it, therefore it's a threat. Similar to negative news, which are fake, and contrary opinions, which are woke propaganda.
I think you are thinking of the wrong 'end'. It isn't talking about the end of the entire movement path, we are talking about the end of the LLMs decision making process, and the output (whether that is the full path the arm should take, or just a subset of the path) is checked against the requirements.
Basically, anything that is leaving the LLM is checked, rather than the internal LLM reasoning process.
> Construction is famously labor intensive: direct labor makes up close to 50% of the cost of constructing a new single-family home in the US, compared to around 6 to 8% of the cost of manufacturing a car.
Is this only calculating last mile labor, or all the labor along the way? Like, for a car, does it count the cost of steel as purely non-labor costs, or do they factor in the labor cost for mining the ore, etc?
Because I feel like at the end of the day, almost all costs are 'labor' costs, they are just paid by different people/companies earlier in the process. Housing just has a lot of the labor as the final step, assembling the house on site. Cars have more of their labor at earlier stages of the process, and the assembling is mostly automated.
As I understand it, even just considering ‘last mile’ costs, labour is a majority of the cost, and fit-out is the majority of the labour. Then again, where I live skilled trades cost a lot, so this is probably different elsewhere.
There were so many scam ads in magazines and newspapers back in the day. Same with all the made for tv stuff and infomercials.
The difference for the newspaper and magazine ads is you couldn't just click on it in an instant. You had to spend a lot more effort and had more chances to realize it is a scam.
In a way, ads like that can be a bit scarier. We recently bought a house and through the next 3 months I got so many scary warnings in the mail, most of which looked legit and official, saying that if I dont visit sign up for X then I will be billed Y, or my house will be at risk of Z, and other such tactics. I was not expecting this inondation but I can see someone falling for this.
Not trying to justify either, BTW. Its just scary how many different ways there are to defraud people.
In other words, provide my AI with tools, instead of providing me an AI that uses your tools.
That way, my AI can bring all the context it needs, and I can bring all of the settings and knowledge about what I want with me. I don't want a fractured world of tons of AIs i interact with where I have to explain all the fundamental information about what I want and how I work every time.
This also has the benefit of sidestepping the issue the essay is talking about. You provide a consistent tool, and the AI weirdness is not your issue anymore. You don't have to worry about solving for all the weird ways people prompt the AI, or the ways they break.
reply