It seems almost too dumb to be true, but apparently the literal belief of SaaS bears is "all software is a 0 because Claude can one-shot these apps"
Just staggering levels of short-sightedness in that statement
Trivedi says the irrational sentiment creates opportunities for patient investors.
It seems almost too dumb to be true, but apparently the literal belief of SaaS bears is "all software is a 0 because Claude can one-shot these apps"
Just staggering levels of short-sightedness in that statement
Some users back Chollet's argument that AI does not make all SaaS worthless and call opposing claims false or simplistic, while others insist SaaS lacks moats and is already or soon will be obsolete.
No Digg Deeper questions have been answered for this story yet.
To be clear:
1. No it can't. I've used Fable while it was available, it was a good model but still less than 1% of the way there.
2. If it could, that fact would generally benefit SaaS companies, not hurt them. Those who stand to benefit most from better developer tools are software developers.
3. The code is not the product. No customer is going to want to take on every side quest they face simply because they can generate code for it, when they could just pay a subscription to get the problem out of their way.
4. Easier code means more software, which means more usage surface for existing SaaS. You're already seeing this trend in the data.
It seems almost too dumb to be true, but apparently the literal belief of SaaS bears is "all software is a 0 because Claude can one-shot these apps"
Just staggering levels of short-sightedness in that statement

@fchollet You’re wrong. Within 6 months the upgraded Fable will be able to 1 shot anything and you won’t have access to it and Anthropic will

@fchollet what do you attribute the never ending mass layoffs of software engineers then?

@fchollet the argument for SaaS=0 is that software will be replaced by agentic execution on demand (not that all software will be bespoke). The abstraction layer of software melts away into a higher level AI execution of the task.
When? That's the question. We are seeing glimpses of it now.

I thought this for a while, but the bottleneck for most large companies is not developer hours, its ideas that work - which extends to saas, bad ideas can be coded just as easily as good ideas now
But the cost of iterations goes down, so I think that benefits high-iteration cultures like startups

@BosonJoe @fchollet Do you have some examples and how your replaced it?

@fchollet coding agents is maybe good for the concept of SaaS, but i think it is bad-to-fatal for the entrenched players in the category whose pattern is to invest heavily in a flagship product... the product-business fit has to have the new production economics in mind from the start

@fchollet so i think the current category leaders (and therefore their indices) will struggle to find direction and pivot, and there will be a new wave of platforms that allow people to solve their own problem in specific domains with 1/5 the effort of vibe-coding their own solution...

@bygregorr @fchollet Platforms for Service Now and Jira would be hard to “one shot”, even with amazing models . SFDC not so much. Same for Dynamics .

@fchollet The truth is somewhere in between. But devs and companies are blind to what indy "devs", with zero experience, can and are producing. I've recently watched an orthodontist produce a softwear masterpiece with Claude, on a whim, with no background or awareness the normal limits.

@fchollet Microwaves are going to put restaurants out of business.

@fchollet and understanding of software in general.
software engineers should be buying up these companies in droves.
@fchollet It's not supposed to be rational, this is how opportunities are created, if one is patient. Although I now only stick to one (and only one) security's chains, and certainly don't play SaaS, I mostly agree with you. Not sure when reality will start to catch up.
It seems almost too dumb to be true, but apparently the literal belief of SaaS bears is "all software is a 0 because Claude can one-shot these apps"
Just staggering levels of short-sightedness in that statement

@fchollet One of the few people who noticed that the code isn't completely dead yet. Respectfully, you're right. Fable isn't really suitable for everything yet, as many say. Basically, there's simply no neural network for everything.

@fchollet How much is big tech paying you to say this

@goodalexander @fchollet The only job in the future will be project management at anthropic

@fchollet not 0 but exponentially less than it used to be

@fchollet Expecting software to be dead, is the opposite of short-sighted. lmao? ur take is the one that is short-sighted. Ur expecting a trend that exists to continue. And a trend that doesn't exist (Ai replacing Saas) to not occur in the future.

@fchollet @levelsio interested on your take on this. I think François has a point here

@goodalexander @fchollet Literally did this with HubSpot in the 72 hours Fable was live. Literally just prompted it to recreate HubSpot with all of its functionality and it just did it.