What happens to SaaS in a world with computer-using agents?

I think this post underestimates how the degree to which “what data is correct” is deeply contextual.<p>My team created an identical hypothesis to this doc ~2 years ago and generated a proof of concept. It was pretty magic, we had fortune 500 execs asking for reports on internal metrics and they’d generate in a couple of minutes. First week we got rave reviews - followed by an immediate round of negative feedback as we realized that ~90% of the reports were deeply wrong.<p>Why were they wrong? It had nothing to do with the LLMs per se, 03-mini doesn’t do much better on our suite than gpt 3.5. The problem was that knowing which data to use for which query was deeply contextual.<p>Digging into use cases you’d fine that for a particular question you needed to not just get all the rows from a column, you needed to do some obscure JOIN ON operation. This fact was only known by 2 data scientists in charge of writing the report. This flavor or problem - data being messy, with the messiness only documented in a few people’s brains, repeated over and over.<p>I still work on AI powered products and I don’t see even a little line of sight on this problem. Everyone’s data is immensely messy and likely to remain so. AI has introduced a number of tools to manage that mess, but so far it appears they’ll need to be exposed via fairly traditional UIs.

This is a good read that is a great starting point for thinking about this. It essentially takes the extreme position - SaaS no longer needs a UI, because the LLM is the UI.<p>In reality, as always, I suspect the truth will be somewhere in between. SaaS products that succeed will be those that have a good UI _and_ and good API that LLMs can use.<p>An LLM is not always the best interface, particularly for data access. For most people, clicking a few times in the right places is preferable to having to type out (or even speak aloud) &quot;Show me all the calls I did today&quot;, waiting for the result, having to follow up with &quot;include the time per call and the expected deal value&quot;, etc etc.<p>There is undoubtedly an opportunity for disruption here, but I think an LLM only SaaS platform is going to be a very tough sell for at least the next decade.

Look, I love LLMs and even implement them for customers, but I am very sceptical about them &#x27;replacing&#x27; ERP and CRP systems. What some AI folks don&#x27;t seem to understand is that traditional ERP and CRP apps are completely driven by auditable business rules because they have to be. If you&#x27;re running a company, there&#x27;s no discretion at all about how money and other assets and liabilities are accounted for. It all has to be strictly according to the rules. This goes for most everything else - management are responsible for the business rules implemented in the system and they need to be precisely spelled out. Sure, AI can and should be used extensively for the human UI piece of it. To simplify getting data into and out of the system for example. But the engine inside and the database are all strictly rule governed and I definitely dont expect that to change anytime soon.

This kind of reminds me of when there was a lot of hype around messenger apps and this idea that we&#x27;d just do everything through a chat interface &#x2F; chat bot.<p>It never panned out, arguably because the technology wasn&#x27;t quite there yet (this was well before ChatGPT came out), but I thought the bigger problem was that people thought that a chat UI was the ultimate user interface. Just didn&#x27;t feel right to me. For simple tasks, sure, but otherwise it felt like for &quot;exploratory&quot; tasks it made more sense to have a graphical user interface of some kind.<p>Same sentiments apply to the hype around agents. Even in a hypothetical world where agents work as well as any human I don&#x27;t think an agent&#x2F;chatbot UI is necessarily the ultimate user interface. If I&#x27;m asking an agent questions, it makes sense for it to show rather than tell in many contexts. Even in a world where agents capture much of the way we interact with computers, it might make more sense for them to show us using 3rd party SaaS apps.

It&#x27;s an intriguing take, but as others have pointed out, the truth will be somewhere in the middle. I don&#x27;t believe that AI will replace the entire SaaS interface. And I also don&#x27;t think it will need as many services and APIs of yester-years.<p>This writeup seems to be authored by a senior designer at Salesforce and I can see the motivation from the their perspective. Their challenges are different than what a new SaaS product will encounter.<p>Like all the incumbents of their time they are a core-ish database that depended on a plethora of point solutions from vendors and partners to fill in the gaps their product left in constructing workflows. If they don&#x27;t take an approach like being discussed here – or in the linked OpenAI&#x2F;Softbank video – they will risk alienating their vendors&#x2F;partners or worse see them becoming competitors in their own right.<p>Disclaimer – I&#x27;m biased too, I&#x27;m building one of the upstarts that aims to compete with Salesforce.

Have you ever watched people talk excitedly about &quot;agents&quot; for thirty or forty years without ever actually providing an example that functioned for more than a couple of very precisely staged demos, if that?<p>You Will.

I think everyone that thinks this way are smoking something. I use the latest and greatest AI tools and they never fail to disappoint, make shit up and just waste hours of time because they would rather answer with nonsense than ask questions or just say I don’t fucking know or something isn’t possible.

I learned about the idea of Generative UI from a Sharp Talk podcast, and it&#x27;s stuck with me ever since.<p>Many SaaS (especially the complex ones, which are the also the most important ones) have a tonne of UI often imposing a huge amount of non-work work onto users - all the clicking you have to do as part of entering or retrieving data, especially if the UI flow doesn&#x27;t fit exactly what you&#x27;re trying to do at that moment. An example might be quicly creating an epic and a bunch of related tickets in Jira, and having them all share some common components.<p>A generative UI would be able to construct a custom UI for the particular thing the user is trying to do at any point in time. I think it&#x27;s a really powerful idea, and it could probably be done today by smartly using eg Jira&#x27;s APIs.<p>The ability to span applications would be even more powerful. Done well it might even kill the need to maintain complex integrations between related Saas (eg how some product development application might need to sync data to&#x2F;from Jira or ADO) by having the AI just keep track of changes and move them from one system to another.<p>Once it gets to the point where the Gen UI is go-to system for interactions you have to wonder what all the designers and UI builders at the myriad SaaS will be doing...

Using SaaS products even with an API is fraught with peril with actual engineers and QA (sometimes) on both sides.<p>Who&#x27;s going to bet millions of dollars these agents after going to get it right. Based on what evidence?

Let&#x27;s look at an actual CRM for a moment. Salesforce has an suite of sales forecasts for projecting sales. A major feature of that is letting people make &quot;adjustments&quot; to the data. Every layer of your sales org can tweak the numbers that the layer below generates: <a href="https:&#x2F;&#x2F;help.salesforce.com&#x2F;s&#x2F;articleView?id=sales.forecasts3_adjustments_overview.htm&amp;type=5" rel="nofollow">https:&#x2F;&#x2F;help.salesforce.com&#x2F;s&#x2F;articleView?id=sales.forecasts...</a><p>I&#x27;m sure some of those adjustments are reasonable, but I&#x27;m also sure this gets used to create a stack of lies to please upper management.<p>There&#x27;s some obvious issues with some sort of AI in such an environment. Do you train the AI to tell the right sorts of lies?

We&#x27;re working on Agents over at Zapier, <a href="https:&#x2F;&#x2F;zapier.com&#x2F;agents" rel="nofollow">https:&#x2F;&#x2F;zapier.com&#x2F;agents</a><p>You can have Agents run behaviors async by attaching triggers to them, for example when you get a specific email or something gets updated in a CRM. You can also give the agent access to basically any third-party action you can think of.<p>Like others in this thread have pointed out, there&#x27;s a nice middle-ground here between an LLM-only interface and some nice UI around it, as well as ways to introduce determinism where it makes sense.<p>The product is still in its early days and we&#x27;re iterating rapidly, but feel free to check it out and give us some feedback. There&#x27;s a decent free plan.

For this to become true, agents first have to transcend &#x27;chatbot&#x27; as the primary interaction layer.<p>There&#x27;s a reason we&#x27;re still using apps instead of talking to Siri…for a huge number of tasks, visual UIs are so much more efficient than long-form text.

I don&#x27;t think the Agents&#x2F;LLM become the UI, they are going to be the orchestrators, but a well though UI is always going to be more useful than having to chat&#x2F;write words so that an agent can help you.<p>It&#x27;s gonna be: reusable saas components + ai orchestrator + specialized UI<p>On a related note, there&#x27;s probably gonna be an extinction level event in the software industry as there&#x27;s no software moat anymore.<p>When every application, every feature, every function can be replicated&#x2F;reproduced by another company in a matter of minutes &#x2F; hours using AI tools, you don&#x27;t have a moat anymore.

This reminds me of the blockchain will make everything obsolete sensation of yesteryear.<p>Why will businesses trust a black box that claims to make good decisions (most of the time) when they have existing human relationships they have vetted, measured, and know the ongoing costs and benefits of?<p>If the reason is humans are expensive, I have news for you. We&#x27;ve had robotics for around 100 years and the humans are still much cheaper than the robots. Adding a bunch of graphics cards and power plants to the mix doesn&#x27;t seem to change that equation in a positive direction.

Continuing on with my &quot;old man yells at cloud&quot; meme of late, here&#x27;s my hot take:<p>So let me get this straight- we are going to train AI models to perform screen recognition of some kind (so it can ascertain layout and detect the &quot;important&quot; ui elements), and additionally ask that AI to OCR all text on the screen so it has some hope of being able to follow some natural language instructions (OCR being a task which, as a HN thread a day or two ago pointed out, AI is <i>exceedingly</i> bad at), and then we&#x27;re going to be able to tell this non-deterministic prediction engine what we want to do with our software, and it&#x27;s just going to do it?<p>Like Homer Simpson&#x27;s button pressing birdie toy? :smackshead:<p>Why do I have reservations about letting a non-deterministic AI agent run my software?<p>Why not expose hooks in some common format for our software to perform common tasks? We could call it an &quot;application programming interface&quot;. We might even insist on some kind of common data interchange format. I hear all the cool people are into EBCDIC nowadays.<p>Then we could build a robust and deterministic tool to automate our workflows. It could even pass structured data between unrelated applications in a secure manner. Then we could be sure that the AI Agent will hit the &quot;save the world&quot; button instead of the &quot;kill all humans&quot; button 100% of the time.<p>On a serious note, we should study various macro recording implementations, to at least have a baseline of what people have been successfully doing for 40+ odd years to automate their workflows, and <i>then</i> come up with an idea that doesn&#x27;t involve investing in a new computer, gpu, and slowly boiling the oceans.<p>This reeks of a solution in search of a problem. And the solution has the added benefit of being inefficient and unreliable. But, people don&#x27;t get billion dollar valuations for macro recorders.<p>Is this what they meant by &quot;worse is better&quot;?<p>Edit: and for the love of FSM, please <i>do not</i> expose any new automation APIs to the network.

So is &quot;AI Agents&quot; something the community has settled on or is this a Google-ism? I remember people arguing about this some time ago with no definitive answer.

In my experience, autonomous tools are not as successful as ones that are built to postulate about and get confirmation of the user’s intent. I think there’s a lot of promise for agents that are built to be controlled by skilled operators.<p>Autonomy is just more sexy, but in my opinion, it’s a poor design direction for a lot of applications.

I wonder how we will train Customer Support to tackle issues faced by LLMs. LLMs can already do basic Customer Support. But stuff like understanding bugs and deciding if they should escalate things to engineers feels like a hard thing for an LLM.

I think the importance to things like user interface, good design are still going to remain just their applications will change to the AI interaction layer &#x2F; control layer which are mentioned in the blog.

SaaS will become the wordpress plugin equivalent for Agent platforms.

UX is already working on this. AI as a first-class persona that can be deliberately designed for and accommodated. APIs and protocols are way too strict. Think HTML and black Times New Roman on white backgrounds from the old days. Clear information (text) and activation options (hyperlink) are all it needs.

Ah yes, we are back to the 90s where we are going to have agents taking care of everything for us. All we are missing is Andersen Consulting to sell this to the CEO.

who has productionized an agent in a setting where there is a low margin for error? I would love to know

Great doc. I wonder when we’ll be getting an OS that dedicates itself to Agents.

We need a simple open-source protocol which includes authentication and ability for agents to make payments. Essentially what you want is the ability for an agent to take a core action (as the article mentions, like adding a record to a CRM).<p>I fundamentally believe that human-oriented web apps are not the answer, and neither is REST. We need something purpose-built.<p>The challenge is, it has to be SIMPLE enough for people to easily implement in one day. And it needs to be open source to avoid the obvious problems with it being a for-profit enterprise.