Ironically the OP is called A_D_E_P_T - ADEPT was the first "proper agent project" I got involved in, back in 1996! I just looked at a couple of deliverables & tidied a demo as it was closing then, but then I did more work on lots of old MAS projects for years after.
Here's the thing - we failed.
Why? Three things.
- service providers like Expedia, Lastminute.com, Zwillow did the data integration by hand. They were able to afford to do this (the bit we didn't expect) because they were able to attract adverts and charge a transaction fee. They had a business model.
- Specific applications like the HMRC self-assessment were able to offer a user experience that gave close control and easy completion. The agents we built were able to do easy completion but they took control off the user and could not perform at 100%. Users were not confident and so didn't use them.
- We did most of the work pre iPhone, so we didn't have the right interface. Of course that has changed... but the Appstore also changed things and user expectations are sky high now.
It was also fersum difficult to program proper MAS. The issue is creating a system based on the expectation of behavior of many other systems which is of course a combinatorial mind bender. We mitigated this a bit with ideas from economics and game theory, but I think we were a long way off solving it.
LLM's do provide extra capability for solving some of these problems, and you young people are so clever and hard working that maybe you can just blitz through some of the others. I can't get at Manus yet, but I don't really expect that it will be a proof of this.
I don’t understand the point of Manus. I don’t need an agent that can kind of do my taxes and kind of plan a vacation and kind of analyze the stock market. If I want to do my taxes, I need a computer program that is very good at doing that and only that. And building that highly specialized computer program looks a lot like writing the software we’ve all been doing for 50 years.
Sure sure sure... or it hallucinates that it sent your tax forms when it didn't. Or it makes an OOPSIE DAISY with some numbers that you overlook. Or or or...
You might not be the target audience or you might not understand the tech yet. These agents are for replacing things like Mechanical Turk and data entry jobs at first.
A couple of years ago, every company added a shitty chat bot to their website. Imagine that without adding a ton of complicated integrations and writing a ton of rules, you can make this chat bot actually become somewhat useful. That would take a ton of work away from customer contact centers.
Imagine that software testing can be as simple as writing: "Create 50 concurrent users that are signing up to my Android app." In stead of writing a shit ton of tests in Robot Framework that need to be updated for every change of the code.
There's some real benefit in AI Agents, even in the short term... But it might not be for end users.
But people are wary of it, they know about the enshitification to come, and it will be horrible beyond comprehension. Those products of the past, have burned right through all layers of trust - and there just is none left, to invest into that dystopian vision of a future where a robot butler wakes you in the morning with advertisement jingles and disregards your shopping list to buy the sponsoring brands.
And all those cooperation, that are supposed to buy AI labor- they know that those mafia companies trying to get in via subscription, will be the worst union of all. Software eating the worlds bread butter indeed.Subscription is up, and its now (costOfEarnings -1$)
People can see the takeover and the faces of it for what they really are and boy is it ugly.
I just want an AI agent that I prompt with my bank account data and the phrase "make me some money", so I can come back after a year and see that bank account doubling.
Meanwhile I'd pretty much spend the time on the beach.
Any significant edge in the market will swiftly be corrected. If it doesn't, the entire market collapses. We can't all sit back at the beach doing nothing while our autonomous AI agents make us all millionaires. Where is that money coming from? Who is generating it?
Money is just an expression of value, either delivered already or promised to be delivered in the future. I can certainly imagine a world where everyone sits back and all the value they require is delivered by robots, they who have better or more robots get more value.
What a nice dream. Unfortunately, we do not live in that world today and a lot of powerful people don't want us to live in that world tomorrow. These people will continue to enforce power structures that let them chiefly benefit from increased automation, not the masses.
In all seriousness if such a system were to exist, why wouldn't the creator simply use it to make great returns for themself instead of scraps for everyone else?
Since you are asking this, there are people, who instead of becoming rich themselves, prefer to sell courses, newsletters and books teaching others how to become rich, out of the goodness of their hearts.
Zvi posted a response[0] to this post, with the following comment:
Ignore the first sentence in Tyler Cowen’s post here, where he asserts that Manus is ‘for real, and ahead of its American counterparts.’ That’s a rather silly way of summarizing the situation, given everything we now know.
> Manus AI is real and ahead of its American counterpart
Isn’t Manus just based on Claude plus browser-use?
I can’t get beyond the first few statement knowing that Manus is just a wrapper around “American counterparts”
It is. They built a non-American AI with American AI and some other parts.
Ironically the OP is called A_D_E_P_T - ADEPT was the first "proper agent project" I got involved in, back in 1996! I just looked at a couple of deliverables & tidied a demo as it was closing then, but then I did more work on lots of old MAS projects for years after.
Here's the thing - we failed.
Why? Three things.
- service providers like Expedia, Lastminute.com, Zwillow did the data integration by hand. They were able to afford to do this (the bit we didn't expect) because they were able to attract adverts and charge a transaction fee. They had a business model.
- Specific applications like the HMRC self-assessment were able to offer a user experience that gave close control and easy completion. The agents we built were able to do easy completion but they took control off the user and could not perform at 100%. Users were not confident and so didn't use them.
- We did most of the work pre iPhone, so we didn't have the right interface. Of course that has changed... but the Appstore also changed things and user expectations are sky high now.
It was also fersum difficult to program proper MAS. The issue is creating a system based on the expectation of behavior of many other systems which is of course a combinatorial mind bender. We mitigated this a bit with ideas from economics and game theory, but I think we were a long way off solving it.
LLM's do provide extra capability for solving some of these problems, and you young people are so clever and hard working that maybe you can just blitz through some of the others. I can't get at Manus yet, but I don't really expect that it will be a proof of this.
I don’t understand the point of Manus. I don’t need an agent that can kind of do my taxes and kind of plan a vacation and kind of analyze the stock market. If I want to do my taxes, I need a computer program that is very good at doing that and only that. And building that highly specialized computer program looks a lot like writing the software we’ve all been doing for 50 years.
I have not looked at Manus but I could imagine that it uses the highly Specialized program for you and fills in the forms based on data it found.
Sure sure sure... or it hallucinates that it sent your tax forms when it didn't. Or it makes an OOPSIE DAISY with some numbers that you overlook. Or or or...
You might not be the target audience or you might not understand the tech yet. These agents are for replacing things like Mechanical Turk and data entry jobs at first.
A couple of years ago, every company added a shitty chat bot to their website. Imagine that without adding a ton of complicated integrations and writing a ton of rules, you can make this chat bot actually become somewhat useful. That would take a ton of work away from customer contact centers.
Imagine that software testing can be as simple as writing: "Create 50 concurrent users that are signing up to my Android app." In stead of writing a shit ton of tests in Robot Framework that need to be updated for every change of the code.
There's some real benefit in AI Agents, even in the short term... But it might not be for end users.
But people are wary of it, they know about the enshitification to come, and it will be horrible beyond comprehension. Those products of the past, have burned right through all layers of trust - and there just is none left, to invest into that dystopian vision of a future where a robot butler wakes you in the morning with advertisement jingles and disregards your shopping list to buy the sponsoring brands.
And all those cooperation, that are supposed to buy AI labor- they know that those mafia companies trying to get in via subscription, will be the worst union of all. Software eating the worlds bread butter indeed.Subscription is up, and its now (costOfEarnings -1$)
People can see the takeover and the faces of it for what they really are and boy is it ugly.
a more perfect union indeed.
I just want an AI agent that I prompt with my bank account data and the phrase "make me some money", so I can come back after a year and see that bank account doubling.
Meanwhile I'd pretty much spend the time on the beach.
Any significant edge in the market will swiftly be corrected. If it doesn't, the entire market collapses. We can't all sit back at the beach doing nothing while our autonomous AI agents make us all millionaires. Where is that money coming from? Who is generating it?
I totally agree with you. I was just trying to be funny.
Money is just an expression of value, either delivered already or promised to be delivered in the future. I can certainly imagine a world where everyone sits back and all the value they require is delivered by robots, they who have better or more robots get more value.
What a nice dream. Unfortunately, we do not live in that world today and a lot of powerful people don't want us to live in that world tomorrow. These people will continue to enforce power structures that let them chiefly benefit from increased automation, not the masses.
"No such thing as a free lunch."
In all seriousness if such a system were to exist, why wouldn't the creator simply use it to make great returns for themself instead of scraps for everyone else?
Since you are asking this, there are people, who instead of becoming rich themselves, prefer to sell courses, newsletters and books teaching others how to become rich, out of the goodness of their hearts.
Also, I think it's existence might disrupt the financial system shomeewhat!
Manus seems to really improve upon existing models by giving them good tooling to use.
Maybe the foundational companies follow suit and make startups like this obsolete?
It makes we wonder now if what we really have is a creativity deficit.
Zvi posted a response[0] to this post, with the following comment:
[0]: https://thezvi.substack.com/p/the-manus-marketing-madnessI saw their posts several times, and thought they were some really shitty spams.
[dead]