I am impressed by how DHH steers clear of bashing the alternative views. He admits that there is value in the other ideas and then frames his perspective and explains why that is valuable.
a renaissance from the olde days we bought the boxed software for hundreds of dollars, for some disks/CDs and a printed manual --- and a oversized box to display on the shelf
Yo guys absolutely love what you do here. I worked for a POS SASS company for 4 years... Made a little dough and bought a small farm way up North by Canada. Took a year off to learn new country boy things. Now I'm getting back into the Dev and needed some fresh vibes. Y'all provide!
I concur with DHH that LLM has similarity to our minds and the way we hallucinate, and similarities to the kindred way they respond to us and impressive in that it can collate and present information at ungodly speed but Jason is also right that we must be mindful what is under the hood (at least today) that it is a word-to-word predictor, it is a compressed mapped-index of known articles and data, it has no volition but to process your inputs held against the system conditions into a line of conversation.
Point being, it has been a mistake to trap data into Slack, Basecamp, in the first place, where making some self-hosted server package is just a standard developer's weekend exercise, and had been even back in 2004. Maybe is great and important that there's now the GNU AGPL to also close the SaaS loophole.
I was going to mention this as well, but found your comment first. Yes, mattermost is a perfectly good open-source alternative to Slack, that you can host yourself. Of course, there are open source IRC servers .. but an easier web front-end seems to be needed for non-IRC-familiar people..
Thank you. I wish the JS ecosystem had someone like DHH over the past decade. I miss his leadership since I moved on from Rails to do isomorphic apps. I love the energy of this conversation. I feel a renaissance in the independent startup ecosystem coming.
This is Nick Carr's "The Big Switch" in reverse. This is de-centralization, which means companies can still can have their own balance sheet. As opposed to the MSFT/AMZN way, which is "You'll own nothing and be happy." I love it.
44:28 “It wasn’t just a road to nowhere, it was a road to hell.” As the saying goes, the road to hell is paved with good intentions. Or another way: the ends justify the means.
Don’t forget Oracle, IBM, Sap etc pre SaaS made fortunes on on premise software ongoing support streams. All saas did is convert high profit margin support contracts to a saas model.
1:12:00 Walking in rough weather can be fun but driving in it can be seriously dangerous. There's a difference between the risk of getting wet or cold and the risk of a fatal car crash. You don't gain a memory of any kind if you don't live through the experience. Walking from my house in harsh weather can be very fun on rare occasions but driving an hour somewhere in it to then walk would be a bunch of very bad stressful driving and potentially lead to a crash.
DHH is always refreshing to watch. But I find Jason Calcanis just insincere. Insincere about his understanding of technology, the markets and the whole ecosystem.
@@GrahamQuigley In this interview specifically his introduction of DHH itself shows he's done just some cursory reading. How can one not be certain that DHH is the creator of the Rails framework especially when he's the only guest. It's 101. Calcanis comes across as a bold font reader. I've followed his work since the Mahalo days and most recently his sycophancy with Musk's Twitter. I know it comes across harsh but that's how he comes across to me.
I actually look forward to Once, although I was underwhelmed by Hey Mail (sorry DHH) looking forward in how the paradigm will shift i suspect - decentralized logic and data, using encryption to protect corp. data on-device - peer-to-peer (for most), light server/proxy to negotiate, - CRDTs - novel version control - allow customers to have escrow/custodianship of the source code [at per-customer level] - multiple device support sounds like a huge revolution
Now...I really connected to this talk...all the way up untill 1:08:56 that they have the audacity to say that there is no more trauma and that your arm could be blown off by a mortar. Well talk to people fleeing jones of conflict. Please make sure you look around when you are making this type of statements. There are right now people that are in that exact situation.
SAAS is good for customers because they pay for what they use. It's good for vendors because their revenue becomes more predictable. I am very skeptical of "post-SAAS".
57:15 those comments about making new episodes are interesting. Even swapping faces in entire movies with people we know could make movies more entertaining to watch. The new software doesn't have to be much more flexible to do some fun and interesting things with entertainment materials.
Happy longtime customer and really enjoying this talk. Also an avid ChatGPT user/subscriber, and I do wonder if Mark Fisher was right in Capitalist Realism and I wonder if AI is capable of propelling us over this hump with truly novel and brilliant visions and ideas... or will it produce mostly small incremental iterations on training data. "Real Art" as we've collectively identified it tends to either be totally ineffable or at least stand-out as extreme examples of the clever interweaving of disparate ideas, often merging distant abstract ideas, mediums, and subject matter in ways that compel us to place enormous value on these instances of "originality". It's almost as if artistic value increases in an inversely correlated way to the popularity or predictability of the art's cultural references.
Post SaaS? This will be news to SaaS companies. The model works fine, but there's a spectrum of value versus cost versus effort to create a product that is really just now settling. Small businesses especially have no desire to own software.
- Embrace AI to enhance productivity and individual capabilities (00:11). - Consider remote work as a cost-effective strategy for startups (00:17). - Explore non-traditional funding and growth methods in the current economic climate (00:23). - Investigate the potential of web apps and software products to reduce operational costs (04:03). - Reevaluate the necessity of large teams and excessive spending in startups (10:25). - Foster independence and resilience in personal and professional life (51:08). - Utilize AI tools like chatbots to streamline tasks and gain quick insights (57:38). - Introduce elements of physical discomfort, like cold showers, to build character and appreciation for comfort (1:13:09). - Remain open to new, challenging, and unconventional ideas to enrich personal perspectives (1:19:01).
Great interview. If I may though, I found myself wanting to hear the end of David's sentences and his points but Jason would interrupt mid sentence. It disrupted the points trying to be made at times and I feel that it may be good to allow the person being interviewed to finish their sentence/thoughts. Not only good for us listening but would give Jason more context then interrupting midway through the sentence.
Interesting conversation, but the form reminds me of American TV - a lengthy “sponsored by / ad” every 10 minutes. Why not make it every 5 minutes? 3 minutes? Mid-sentence, even? Like David says, Tiktok time spans. 🤦🏻♂
Remote work is fine but if it’s an excuse to exploit people . While the work is just as hard and require sacrifices ( sedentary lifestyle ) , training all the time ( to be better communicator to be better this or that ) and you can’t even get a living wage compared to all the hours you put in . I will start to say to young people to do something else . And myself too. What’s the point? If only the ceo can afford to have a house and a family, is that a viable future for the world. I understand that in capitalism self interest is the driving force. Well self interest of the next generations should be to do something else. Because be a real IT SWE with deep knowledge requires a lot of time. If you can’t even be paid for that. What’s the point if you can’t even have a life after work… nah people will either start slacking or put less intensity. I really believe in “you get out what you put in”
What’s problematic with AI is the name itself. Implies intelligence when it’s a work in progress. As a UX designer who is excited by the ability to eliminate redundant workflows for users with “AI” and its associated tech (nlp, ocr etc) I fear the point when we begin marketing AI as doing or providing something that may not be accurate. It’s not the tech itself but the way we respond to it. And as we have seen with social media not all users play with their toys in the same way and their are consequences to our society. Let’s not stop having these conversations as the technology evolves and is adopted! 🎉
Definitely can relate. When we grew, we had the same challenges too. I was very naive. Maybe still am 😅 Now, we just do more with less with lots of tech and AI. P/s Once resonates with us. Lots of subscriptions are so painfully a waste of hard earned $
People pay saas to not be bothered with hosting and maintenance. there are open source alternative to almost everything. paying millions on a chat is dumb though
the thing is developoers outside of SF also want that quarter of a million as a sallary, it might be a misguided want because that barely allows them to live in bay area, but still we all want it.
I love how Jason’s personal example of childhood adversity is going skiing 🤢 - or rather 💩 Enjoyed the tech parts of the discussion but the segment on moral character I could do without
the irony is that all the entrepreneurs are so focused on the MVPs they forgot to make a Minimum Viable Company because that is also part of The Lean Startup Method. Everyone has been on the hurry to grow and sell.
he made millions off saas model after he's been paying millions to aws and slack and other saas companies he realized now i'm going to double down on pay-once model, genius.
There are times in history for ideas. That’s what he talked about. SaaS was excellent in its time and perhaps now it doesn’t make as much sense for certain segments of the market.
He is right I have to basically not raise money until much better times. It's nice to focus on product instead of investors I feel like since co s coming out of this time will be very strong and profitable because of the harsh journey
The problem with SAAS is that it gets real expensive. The VCs all want to make a pile. There needs and will be lots of competition in the software space.
I'm thinking politicians candidates should run on making WFH/remote work a worker's right, equivalent to the accommodations that companies have to do for people with different capacities. Now, I love WFH because saves time to everyone, and frees all the city services from overload so then only those that actually need it will be using it. Management's rejection of WFH is just a failure of leadership a skill issue in the leader to learn how to read its team efforts via remote activity.
OK, but there's this problem of a contradiction: 53:17 the facts are wrong because it's just a generator/permutator for whatever correct or junk data some people put onto the WWW. But in the subsequent exchange, it's the greatest tutor explaining Einstein and the Israel-Palestine conflict, and such an "acceleration"? Maybe this is because image variation might be more useful (artistically) than text generation (except in marketing/advertising/copywriting, generating homework, research papers, social media posts). 54:11 Human performance is very poor (think their memory or performing math), that's why a machine that would do it better, properly, be correct and not fabricate, would be far more desirable than falling for a new version of ELIZA.,
I don't always agree with DHH, in fact sometimes i think he's drunk, but he do get a lot of things right. Which as it all seems it goes to his favor in total.
Agreed with a lot of this until the logo stuff came about; not sure having AI effectively plagiarize work for others benefit whilst simultaneously harming the job market is a good thing.
But hey, Jason’s gonna enjoy a remix of a remix of a remix of Dire Straits and the n-th custom re-run of The Sopranos til his death, he’s so happy about it… He’s now A Prompt Artist with a Stratocaster blurred in the background. Everyone gets a logo and a 4-chord progression, only $19,99 a month, till GPUs last.
That part outlined everything I hate about AI. I feel like AI should be used to simplify workflows so that there’s MORE creative time to create genuine works of art, not computer generated schlock with no soul 😒
@@TheRonnieaj - We haven’t built AI (yet). It’s an acronym that is easier to pronounce and understand than Statistical Predictive LLM using a multi-head attention mechanism or something even longer. In other words, it's not intelligent, but indeed generative of whatever quantity of “soulless schlock” (to use your term) the engineers have stolen from the human creators (without compensating them). LLMs are probably best at one thing - as a mirror of most of the humanity. To be truly useful and interesting, everything that the current systems spew out require the “taste” and “discernment” and even “technical skill” that most of 8B on the planet do not possess, but continue to dream of. That’s what Jason says openly. He could never play like Mark Knopfler nor draw like “the illustrator we used to pay 5000 USD for a month of work” and relishes in the fact that that horrible DallE logo “is better, faster”. No, it is not better. I’d be so ashamed of such bad taste. I’d keep quiet. But hey, I _can_ and have drawn a lot of logos and create/arrange/master my own music. Just different intellectual planes, non-intersecting, except at the points of “hey, we brute-forced statistics to hallucinate some of your 'magical' stuff, you snobbish creators, here’s our revenge”. I suspect the truly thinking / creative AI will laugh at such statements. A laughter that won’t be understood.
the original web was just too powerful that's why they had to lock it down and use custom app stores for native app distribution instead of using the browser for native app distribution. the original web was a interconnected file-system where you can be a folder on it (extreme freedom and power)!
I think DHH is right until a certain point. He misses the mark about Slack. No company wants to manage Slack internally. Yes, it costs money, but it still costs less than having to manage it internally while risking to shut down developers when things go wrong. Some SaaS will die, but most will thrive, because they bring value, they are fast to setup, and most importantly, the company doesn't have to worry about infra and having an expert-on-call to solve problems when things go wrong.
Jury is very much out on Twitter. It's a big assumption that the changes are sustainable. Social media is less about technical features and more about moderation and community features in the age of polarisation. You could argue with the rebranding, change in focus and ethos, twitter is already dead, gone the way of blockbuster, blackberry and Nokia. Not a shining example of cutting excess fat and thriving.
Noone likes to install software and manage servers, come on.
2 місяці тому
I agree with most things DHH said in this interview, but I am a bit unsure about his take on AI. It's not always good to be an optimist. Doomerism isn't the best approach, either. Let's follow the middle path represented by AI safety research. Most AI safety researchers are not against AI, they simply want to make it safe in the long term. AI shouldn't be like the internet or cryptography, as DHH suggests. It should be like airplanes, rockets, or nuclear power plants -- we should innovate, but also have high safety standards. That said, when AI safety researchers discuss the dangers of AI, they don't really mean ChatGPT and DALL-E. These technologies are wonderful, but they are not powerful enough to pose an existential risk.
DHH is great. Feedback to Jason Calacanis: imho you should improve your listening skills and not interrupt the person being interviewed that much. It's annoying
he is basically saying nothing just to justify his new venture. the internet goes down frequently in the 90s and early 2000s. There were many weak links and plenty of last mile failures. Simply put, there are just some products that are so easy to run and basically is one machine operation. there is nothing disruptive by a stretch about this.
A lot smaller sized companies will come in with buy once products thst match expensive SaaS products. Tech will move from SaaS to company self hosted infrastructure again.
I am impressed by how DHH steers clear of bashing the alternative views. He admits that there is value in the other ideas and then frames his perspective and explains why that is valuable.
a renaissance
from the olde days we bought the boxed software for hundreds of dollars, for some disks/CDs and a printed manual --- and a oversized box to display on the shelf
Wow, what a guest and what an execution of this interview Jason. Wonderful content.
I appreciate DHH's passion for what he does. I think he has more great work ahead of him.
Yo guys absolutely love what you do here. I worked for a POS SASS company for 4 years... Made a little dough and bought a small farm way up North by Canada. Took a year off to learn new country boy things.
Now I'm getting back into the Dev and needed some fresh vibes. Y'all provide!
Thank you and good luck!
That's awesome. Love your music!
Always love hearing David speak.
DHH is the man
This was amazing. Thank you, Jason, DHH and TWiSt.
10/10. Thank you both for your candor!
So excited to hear this. As a rails developer and big fan of DHH. He has such great insights❤
I concur with DHH that LLM has similarity to our minds and the way we hallucinate, and similarities to the kindred way they respond to us
and impressive in that it can collate and present information at ungodly speed
but Jason is also right that we must be mindful what is under the hood (at least today) that it is a word-to-word predictor, it is a compressed mapped-index of known articles and data, it has no volition but to process your inputs held against the system conditions into a line of conversation.
Amen 🙏
It was such a pleasure an such a great episode. Thank you guys for creating this.
Hey Jason, you can export all your slack users and history using its API and import it all into a self-hosted open source MatterMost server! 😊
Point being, it has been a mistake to trap data into Slack, Basecamp, in the first place, where making some self-hosted server package is just a standard developer's weekend exercise, and had been even back in 2004. Maybe is great and important that there's now the GNU AGPL to also close the SaaS loophole.
I was going to mention this as well, but found your comment first. Yes, mattermost is a perfectly good open-source alternative to Slack, that you can host yourself.
Of course, there are open source IRC servers .. but an easier web front-end seems to be needed for non-IRC-familiar people..
Great convo, and great perspective from DHH.
That said, can't say I relate to considering going on a ski trip as "adversity".
Yeah, as DHH said, every generation of old men has to say how tough they had it, I guess that applies to going skiing and hiking for some people.
This was an amazing episode. DHH is valuable to listen to. Thank you
Thank you. I wish the JS ecosystem had someone like DHH over the past decade. I miss his leadership since I moved on from Rails to do isomorphic apps. I love the energy of this conversation. I feel a renaissance in the independent startup ecosystem coming.
This is Nick Carr's "The Big Switch" in reverse. This is de-centralization, which means companies can still can have their own balance sheet. As opposed to the MSFT/AMZN way, which is "You'll own nothing and be happy." I love it.
Super interesting conversation. Thank you!
Glad you enjoyed it!
I love this interview.
Such a great perspective.
David just spraying knowledge! What a great discussion! Thank you both for this.
I wish this had more mid-episode commercials
44:28 “It wasn’t just a road to nowhere, it was a road to hell.”
As the saying goes, the road to hell is paved with good intentions. Or another way: the ends justify the means.
"The future is already here, it's just not widely distributed" is so true.
Before slack, we used IRC in a world long long ago, far far away, for free.
Don’t forget Oracle, IBM, Sap etc pre SaaS made fortunes on on premise software ongoing support streams. All saas did is convert high profit margin support contracts to a saas model.
Finally, someone that makes a hell of a lot of sense 🙏🏻
Rocket Chat is a great opensource self host Slack alternative
Very good podcast, no BS perspective 👍
Holy crap I’ve never been in a channel with so many ads.
1:12:00 Walking in rough weather can be fun but driving in it can be seriously dangerous. There's a difference between the risk of getting wet or cold and the risk of a fatal car crash. You don't gain a memory of any kind if you don't live through the experience. Walking from my house in harsh weather can be very fun on rare occasions but driving an hour somewhere in it to then walk would be a bunch of very bad stressful driving and potentially lead to a crash.
DHH is always refreshing to watch. But I find Jason Calcanis just insincere. Insincere about his understanding of technology, the markets and the whole ecosystem.
How do you experience him as being insincere?
@@GrahamQuigley In this interview specifically his introduction of DHH itself shows he's done just some cursory reading. How can one not be certain that DHH is the creator of the Rails framework especially when he's the only guest. It's 101. Calcanis comes across as a bold font reader. I've followed his work since the Mahalo days and most recently his sycophancy with Musk's Twitter. I know it comes across harsh but that's how he comes across to me.
What's the business model in terms of providing long term support for the software? Will there be ongoing bug fixes and security updates?
I actually look forward to Once, although I was underwhelmed by Hey Mail (sorry DHH)
looking forward in how the paradigm will shift
i suspect
- decentralized logic and data, using encryption to protect corp. data on-device
- peer-to-peer (for most), light server/proxy to negotiate,
- CRDTs
- novel version control
- allow customers to have escrow/custodianship of the source code
[at per-customer level]
- multiple device support
sounds like a huge revolution
@boonkiathan you dont answer the question
There is none, it's a flawed idea and it hasn't been thought through. He needs to make money off his once domain lmao
Now...I really connected to this talk...all the way up untill 1:08:56 that they have the audacity to say that there is no more trauma and that your arm could be blown off by a mortar. Well talk to people fleeing jones of conflict. Please make sure you look around when you are making this type of statements. There are right now people that are in that exact situation.
"My son has a gang...DALLE helped him make t-shirts..." lol. Great podcast!
It was a delicious conversation. Accelerate!
SAAS is good for customers because they pay for what they use. It's good for vendors because their revenue becomes more predictable. I am very skeptical of "post-SAAS".
Great point about what if your top 30% leaves when return to office is implemented.
LOL, pay once until the next security patch. There's a reason even MSFT doesn't sell packaged apps anymore.
57:15 those comments about making new episodes are interesting. Even swapping faces in entire movies with people we know could make movies more entertaining to watch. The new software doesn't have to be much more flexible to do some fun and interesting things with entertainment materials.
Thank you, Gentlemen. I was losing hope, now I see a way forward.
Happy longtime customer and really enjoying this talk. Also an avid ChatGPT user/subscriber, and I do wonder if Mark Fisher was right in Capitalist Realism and I wonder if AI is capable of propelling us over this hump with truly novel and brilliant visions and ideas... or will it produce mostly small incremental iterations on training data. "Real Art" as we've collectively identified it tends to either be totally ineffable or at least stand-out as extreme examples of the clever interweaving of disparate ideas, often merging distant abstract ideas, mediums, and subject matter in ways that compel us to place enormous value on these instances of "originality". It's almost as if artistic value increases in an inversely correlated way to the popularity or predictability of the art's cultural references.
Big fan of Ruby on Rails!
Don't know how I found this but Hell Yeah!
Post SaaS? This will be news to SaaS companies. The model works fine, but there's a spectrum of value versus cost versus effort to create a product that is really just now settling. Small businesses especially have no desire to own software.
I wonder if Jcal will still call the style transfer art generation "democratisation", even after the openai lawsuit?
- Embrace AI to enhance productivity and individual capabilities (00:11).
- Consider remote work as a cost-effective strategy for startups (00:17).
- Explore non-traditional funding and growth methods in the current economic climate (00:23).
- Investigate the potential of web apps and software products to reduce operational costs (04:03).
- Reevaluate the necessity of large teams and excessive spending in startups (10:25).
- Foster independence and resilience in personal and professional life (51:08).
- Utilize AI tools like chatbots to streamline tasks and gain quick insights (57:38).
- Introduce elements of physical discomfort, like cold showers, to build character and appreciation for comfort (1:13:09).
- Remain open to new, challenging, and unconventional ideas to enrich personal perspectives (1:19:01).
Great interview. If I may though, I found myself wanting to hear the end of David's sentences and his points but Jason would interrupt mid sentence. It disrupted the points trying to be made at times and I feel that it may be good to allow the person being interviewed to finish their sentence/thoughts. Not only good for us listening but would give Jason more context then interrupting midway through the sentence.
We finally have someone matching or even exceeding Jason's energy :)
First 30 minutes were ok, the rest felt like a rant.
Good stuff
Interesting conversation, but the form reminds me of American TV - a lengthy “sponsored by / ad” every 10 minutes. Why not make it every 5 minutes? 3 minutes? Mid-sentence, even? Like David says, Tiktok time spans. 🤦🏻♂
Great interview!!
Isn’t this exactly what Mattermost is?
Remote work is fine but if it’s an excuse to exploit people . While the work is just as hard and require sacrifices ( sedentary lifestyle ) , training all the time ( to be better communicator to be better this or that ) and you can’t even get a living wage compared to all the hours you put in . I will start to say to young people to do something else . And myself too. What’s the point? If only the ceo can afford to have a house and a family, is that a viable future for the world. I understand that in capitalism self interest is the driving force. Well self interest of the next generations should be to do something else. Because be a real IT SWE with deep knowledge requires a lot of time. If you can’t even be paid for that. What’s the point if you can’t even have a life after work… nah people will either start slacking or put less intensity. I really believe in “you get out what you put in”
Been remote since 2007 with a few short years in between. The office is a joke.
What’s problematic with AI is the name itself. Implies intelligence when it’s a work in progress. As a UX designer who is excited by the ability to eliminate redundant workflows for users with “AI” and its associated tech (nlp, ocr etc) I fear the point when we begin marketing AI as doing or providing something that may not be accurate. It’s not the tech itself but the way we respond to it. And as we have seen with social media not all users play with their toys in the same way and their are consequences to our society. Let’s not stop having these conversations as the technology evolves and is adopted! 🎉
Is mattermost not valid slack alternative?
TOO MANY garbage ads. Just slim it down will you?!!!!
Definitely can relate. When we grew, we had the same challenges too. I was very naive. Maybe still am 😅 Now, we just do more with less with lots of tech and AI. P/s Once resonates with us. Lots of subscriptions are so painfully a waste of hard earned $
People pay saas to not be bothered with hosting and maintenance. there are open source alternative to almost everything. paying millions on a chat is dumb though
Great convo
Thank God someone is coming to their senses.
the thing is developoers outside of SF also want that quarter of a million as a sallary, it might be a misguided want because that barely allows them to live in bay area, but still we all want it.
I love how Jason’s personal example of childhood adversity is going skiing 🤢 - or rather 💩
Enjoyed the tech parts of the discussion but the segment on moral character I could do without
the irony is that all the entrepreneurs are so focused on the MVPs they forgot to make a Minimum Viable Company because that is also part of The Lean Startup Method. Everyone has been on the hurry to grow and sell.
42:00 they are affectionately known as "The Cathedral"
Agree with all the SaaS out there today that really shouldn't be. It's a business incentive, not a customer benefit.
he made millions off saas model after he's been paying millions to aws and slack and other saas companies he realized now i'm going to double down on pay-once model, genius.
😂
There are times in history for ideas. That’s what he talked about.
SaaS was excellent in its time and perhaps now it doesn’t make as much sense for certain segments of the market.
Wow. This got me thinking… what ONCE business model can bite off of?…
He is right I have to basically not raise money until much better times. It's nice to focus on product instead of investors I feel like since co s coming out of this time will be very strong and profitable because of the harsh journey
What to do about piracy? SaaS prevents that.
Will future versions / upgrades of a "once" product be free too? That would be nice.
Doesn't make sense. Resharper is a great version of a once product
31:41 thats me. But Im realizing most of these businesses also dont need to either. They just want the cushy-ness of working on a startup.
The problem with SAAS is that it gets real expensive. The VCs all want to make a pile. There needs and will be lots of competition in the software space.
I'm thinking politicians candidates should run on making WFH/remote work a worker's right, equivalent to the accommodations that companies have to do for people with different capacities.
Now, I love WFH because saves time to everyone, and frees all the city services from overload so then only those that actually need it will be using it.
Management's rejection of WFH is just a failure of leadership a skill issue in the leader to learn how to read its team efforts via remote activity.
OK, but there's this problem of a contradiction: 53:17 the facts are wrong because it's just a generator/permutator for whatever correct or junk data some people put onto the WWW. But in the subsequent exchange, it's the greatest tutor explaining Einstein and the Israel-Palestine conflict, and such an "acceleration"? Maybe this is because image variation might be more useful (artistically) than text generation (except in marketing/advertising/copywriting, generating homework, research papers, social media posts). 54:11 Human performance is very poor (think their memory or performing math), that's why a machine that would do it better, properly, be correct and not fabricate, would be far more desirable than falling for a new version of ELIZA.,
I don't always agree with DHH, in fact sometimes i think he's drunk, but he do get a lot of things right.
Which as it all seems it goes to his favor in total.
Agreed with a lot of this until the logo stuff came about; not sure having AI effectively plagiarize work for others benefit whilst simultaneously harming the job market is a good thing.
But hey, Jason’s gonna enjoy a remix of a remix of a remix of Dire Straits and the n-th custom re-run of The Sopranos til his death, he’s so happy about it… He’s now A Prompt Artist with a Stratocaster blurred in the background.
Everyone gets a logo and a 4-chord progression, only $19,99 a month, till GPUs last.
That part outlined everything I hate about AI. I feel like AI should be used to simplify workflows so that there’s MORE creative time to create genuine works of art, not computer generated schlock with no soul 😒
@@TheRonnieaj - We haven’t built AI (yet). It’s an acronym that is easier to pronounce and understand than Statistical Predictive LLM using a multi-head attention mechanism or something even longer.
In other words, it's not intelligent, but indeed generative of whatever quantity of “soulless schlock” (to use your term) the engineers have stolen from the human creators (without compensating them).
LLMs are probably best at one thing - as a mirror of most of the humanity.
To be truly useful and interesting, everything that the current systems spew out require the “taste” and “discernment” and even “technical skill” that most of 8B on the planet do not possess, but continue to dream of.
That’s what Jason says openly. He could never play like Mark Knopfler nor draw like “the illustrator we used to pay 5000 USD for a month of work” and relishes in the fact that that horrible DallE logo “is better, faster”.
No, it is not better. I’d be so ashamed of such bad taste. I’d keep quiet. But hey, I _can_ and have drawn a lot of logos and create/arrange/master my own music.
Just different intellectual planes, non-intersecting, except at the points of “hey, we brute-forced statistics to hallucinate some of your 'magical' stuff, you snobbish creators, here’s our revenge”.
I suspect the truly thinking / creative AI will laugh at such statements. A laughter that won’t be understood.
the original web was just too powerful that's why they had to lock it down and use custom app stores for native app distribution instead of using the browser for native app distribution. the original web was a interconnected file-system where you can be a folder on it (extreme freedom and power)!
Jason - it's call E/ACC and no-one runs it.
This was more This Week in Sociology.
I think DHH is right until a certain point. He misses the mark about Slack. No company wants to manage Slack internally. Yes, it costs money, but it still costs less than having to manage it internally while risking to shut down developers when things go wrong. Some SaaS will die, but most will thrive, because they bring value, they are fast to setup, and most importantly, the company doesn't have to worry about infra and having an expert-on-call to solve problems when things go wrong.
DESKTOP is back!!!!
Why does it read Post ass era on the thumbnail? 🤔
It's 1996
So DHH advocates for deploying our apps on Apple laptops instead of AWS?
Why not?
DHH! Lets Goo
Jury is very much out on Twitter. It's a big assumption that the changes are sustainable. Social media is less about technical features and more about moderation and community features in the age of polarisation. You could argue with the rebranding, change in focus and ethos, twitter is already dead, gone the way of blockbuster, blackberry and Nokia. Not a shining example of cutting excess fat and thriving.
love DHH!!!
Noone likes to install software and manage servers, come on.
I agree with most things DHH said in this interview, but I am a bit unsure about his take on AI. It's not always good to be an optimist. Doomerism isn't the best approach, either. Let's follow the middle path represented by AI safety research. Most AI safety researchers are not against AI, they simply want to make it safe in the long term. AI shouldn't be like the internet or cryptography, as DHH suggests. It should be like airplanes, rockets, or nuclear power plants -- we should innovate, but also have high safety standards. That said, when AI safety researchers discuss the dangers of AI, they don't really mean ChatGPT and DALL-E. These technologies are wonderful, but they are not powerful enough to pose an existential risk.
This time is different with AI
What’s the business model then? Sell slack for $1000?
It’s a lot of blahblah so far.
I like that model. 😂
@@unitedwindows6884 you’ll pay the same amount in service
I miss software products
Of all people to listen to DHH chose Thomas sowell, 🤦🏾🤦🏾
Seeing this for the first time and For a moment i felt like i was looking at “Christ” talking about SAAS 😀
Funny I come from the SAP world and here we go in the exact opposite direction
DHH is great. Feedback to Jason Calacanis: imho you should improve your listening skills and not interrupt the person being interviewed that much. It's annoying
"Man invents fictional scenario and then gets angry about it." 😊
he is basically saying nothing just to justify his new venture. the internet goes down frequently in the 90s and early 2000s. There were many weak links and plenty of last mile failures. Simply put, there are just some products that are so easy to run and basically is one machine operation. there is nothing disruptive by a stretch about this.
A lot smaller sized companies will come in with buy once products thst match expensive SaaS products. Tech will move from SaaS to company self hosted infrastructure again.
I read this as SASS is dead 😂, I was like what's wrong with SASS?
Dude lost me when he became another old dude yelling about woke culture
Once is aptly named as the future isn't capitalist.