“Vaughan was surprised to find it was often the technical staff, not marketing or sales, who dug in their heels.”
So the people that understood it best were sceptical, and this didn’t give him pause.
Can someone explain to me why all these empty suits dick ride LLMs so hard?
Because they try the tools, realize that their job is pretty much covered by LLMs and think it’s the same for everyone.
They’re easily conned and they love yes men.
Technical staff were skeptical because they actually know what AI can and can’t do reliably in production environments - it’s good at generating content but terrible at logical reasoning and mission-critical tasks that require consistancy.
So it’s the CEO they should replace.
it’s good at generating content but terrible at logical reasoning and mission-critical tasks that require consistency.
Thank goodness nobody is crusading to have AI take over medicine.
…which is why I categorically refuse to use the term Artificial intelligence .
The bullshitters were quick to adopt the bullshit factory.
Can someone explain to me why all these empty suits dick ride LLMs so hard?
$$$$$$$
AIs are cheaper than humans.
Cheaper NOW, when OpenAI operates at huge loss
Not really. Because they don’t work and then they have to hire more humans which is more expensive than just keeping them on for the 6 months it’ll take for the CEOs to realise that.
Today, I ran into a bug. We’re being encouraged to use AI more so I asked copilot why it failed. I asked without really looking at the code. I tried multiple times and all AI could say was ‘yep it shouldn’t do that’ but didn’t tell me why. So, gave up on copilot and looked at the code. It took me less than a minute to find the problem.
It was a switch statement and the case statement had (not real values) what basically reads as ’ variable’ == ‘caseA’ or ‘caseB’. Which will return true… Which is the bug. Like I’m stripping a bunch of stuff away but co-pilot couldn’t figure out that the case statement was bad.
AI is quickly becoming the biggest red flag. Fast slop, is still slop.
AI thinks in the same way that ants think, there’s no real intelligence or thought going on but ants are still able to build complex logistics chains by following simple rules, although AI works on completely different principles the effect is the same, it’s following a lot of simple rules that lead to something that looks like intelligence.
The problem is a lot of people seem to think that AIs are genuinely simulations of a brain, they think the AI is genuinely conjugating because they kind of look like they do sometimes. The world is never going to get taken over by a mindless zombie AI. If we ever do get AGI it won’t be from LLMs that’s for sure.
I do find AI useful when I’m debugging a large SQL / Python script though and gotta say I make use of it in that case… other than that it’s useless and relying on it as ones main tool is idiotic
“Vaughan was surprised to find it was often the technical staff…”
Tell me you’re completely out of touch with your company and what it does without telling me you’re completely out of touch with your company and what it does. FFS how is this guy the CEO? Oh, he’s one of the founders? Brilliant.
Vaughan says he didn’t want to force anyone. “You can’t compel people to change, especially if they don’t believe.”
But he did. Change or be fired, basically.
“You multiply people…give people the ability to multiply themselves and do things at a pace,” he said, touting the company’s ability to build new customer-ready products in as little as four days, an unthinkable timeline in the old regime.
Ooh I bet some nefarious hacker types will be salivating at the incredibly rushed code base that is probably a spaghetti mess and as insecure as fuck.
Vaughan disclosed that the company, which he said is in the nine-figure revenue range, finished 2024 at “near 75% Ebitda”—all while completing a major acquisition, Khoros.
I had to look up EBITDA - some interesting points to consider when you look at this metric he used:
A negative EBITDA indicates that a business has fundamental problems with profitability. A positive EBITDA, on the other hand, does not necessarily mean that the business generates cash. This is because the cash generation of a business depends on capital expenditures (needed to replace assets that have broken down), taxes, interest and movements in working capital as well as on EBITDA.
While being a useful metric, one should not rely on EBITDA alone when assessing the performance of a company. The biggest criticism of using EBITDA as a measure to assess company performance is that it ignores the need for capital expenditures in its assessment.Hmmm… I’m no accountant (I leave that to my actual accountant), but surely if they were being profitable it would sound better to say something like “We’ve remained profitable throughout and our earnings per quarter are on par if not greater than before.”?
I’m no accountant but surely if they were being profitable it would sound better to say something like “We’ve remained profitable throughout and our earnings per quarter are on par if not greater than before.”?
no, because profitability isn’t the key figure they are interested in. it’s growth. i recently got fired because of disappointing growth; e.g. the increase in profitability was not as large as they expected. which means they still made more money than last year.
this is why expenditures get relegated to “externality” status; because otherwise projections would make it look like a company can not grow infinitely large, and surely that’s not true
I’ve never heard of this jackass nor his shitty software. I feel privileged.
I wonder if he thinks we’re dumb or just doesn’t care. They’d have been laid off either way. “Return to work”, “Stack ranking”, “AI refusal”, whatever you say bro.
“It enabled us to shit out products in 4 days.”
Glad they incorporated such thorough testing in their process.
One guy is like “Friday is forced AI ‘training’ day” (as if one must ‘train’ to write prompts. Using natural language rather than a unique language or syntax and trusting the computer to make a comprehensible and accurate output is the whole point), and then he has the gall to claim “turns out people hate learning!”
Writing prompts is definitely a thing users must learn to do properly, to get the right results.
But anyways, any company that fires people in favor of AI is only digging their own grave anyways. I personally believe AI (of which LLM is only a small part) can definitely serve as an automation tool that can increase output. Great companies will use this tech to give their employees more time to work on things that are meaningful to the company, that the AI cannot do. For instance, a company could free up some time of highly skilled engineers to help a couple hours a week on the most complicated service desk issues to increase customer satisfaction. Or the LLM can create more time for sales to have meetings with customers, instead of doing admin they already hate, etc… Use it to grow, not to shrink.
Besides, if your company can be completely run by AI anyways, then congratulations, you just reached the end goal of open sourcing your company. Because why the heck won’t anyone be able to replicate that quickly?
Besides, if your company can be completely run by AI anyways, then congratulations, you just reached the end goal of open sourcing your company. Because why the heck won’t anyone be able to replicate that quickly?
Yeah that’s the thing these tech bros never seem to understand. It’s obviously not going to work because if it did work it would have already been done by somebody else, it’s called the Law Of Mediocrity. It’s simply requires the base assumption that you are not the smartest person in the universe, which of course is where it all falls down, because they always assume they are.
One little thing AI can’t do is probably the reason why I also use AI with caution. I use it for all the bullshit emails and communication I have to keep doing just to stay employed. But there’s this one little trick it can’t do. Sure it can summarize a resume or a book or give me the equation to calculated the size of Pythagora’ss triangular dick. But the one little thing it really can’t do is thinking. AI can’t think and come up with original content. It can only mimic and regurgitate old ideas and thoughts, not new ones.
Many people can’t come up wit new ideas either 😜
I may not have a lot of respect for some of my co-workers, and frankly a lump of lard would be an improvement, bit even the most useless human can out think an AI when it comes to anything slightly out of the box.
My nephew got in a bit of trouble at school a while ago because he answered a question “write a sentence containing the word 'why”." With “Why?” You can ask an AI dozen times the same thing and it will always just do the obvious thing, it’ll never be original. He’s six and he can outthink an AI.
You took my joke too literally
Very true if not real reality at its finest.
You can ask it to synthesis information. More specially you can ask it to compare the data and ask to give examples of other things that share the same attributes as the other 2. Also, if I’m painting I sometimes ask about color, but I just got a color wheel.
CEO of enterprise-software powerhouse IgniteTech.
Can someone tell me what they do? They don’t have a Wikipedia Article and their website is mostly AI slop.
They throw buzzwords at venture capitalists in hopes of one day selling out.
After grilling their silly LLM for a while, I was able to squeeze out what that company really is all about. They don’t really make anything. They just buy miscellaneous software companies, and turn those apps into subscription based cloud cancer. Enterprise software meets maximum enshittification, yeah baby!
Ah, so removing employees from this dumpster fire was a net positive for society.
I think only bankruptcy is the net positive, as long as they don’t stiff legitimate creditors.
No don’t you see - fewer employees means there’s less of anything getting done, and this company is just a parasite that produces nothing of value.
They throw buzzwords
Now I understands why the CEO thinks AI could replace everybody.
deleted by creator
I understand what enterprise software is. That wasn’t my question.
“Doing Our Part to Make the World a Greener Place”
Clown company. You can’t promote AI and do a claim like that at the same time.
By accelerating the collapse of a human survivable ecosystem we will bring about the end of humanity, resulting in a greener environment for the handful of surviving species.
By
accelerating the collapse ofpivoting a human survivable ecosystem we willbring about the endaccelerate a paradigm shift of humanity, resulting in a greener environment for thehandfulstable base ofsurviving species.recurring revenue.
Just like an AI. Instead of learning from mistakes, he repeates them, and denies any wrongdoing.
“You’re Absolutely right!”
Because he asks the ai what’s best but the chatbot always treats it as a loaded question and it wants to be seen as helpful so it finds a way to agree yes-man style.
“The marketing and salespeople were enthused by the possibilities of working with these new tools, he added.”
https://youtu.be/KHJbSvidohg#t=13s
I see the same push where I work and I cannot get a good answer to the most basic question:
“Why?”
“We want more people using AI.”
“Why?”
“. . .”
I usually ignore these kind of trends. Just meet any required deadlines etc but don’t engage too much. The vast majority will just disappear.
Specifically as a software developer I cannot see a good outcome from engaging with this trend either. It’s going to go one of two ways.
1: It pans out sooner rather than later that AI wasn’t the panacea they thought it was, and it either is forgotten about, or becomes a set of realized tools we use, but don’t rely on.
2: They believe it can replace us all, and so they replace us all with freshly graduated vibe “programmers” and I don’t have a job anyway.
I don’t really see an upside to engaging with this in any kind of long term plan.
2. It’s about breaking the power of tech workers by reducing them from highly skilled specialists to interchangeable low-status workers whose job is to clean up botshit until it compiles. (Given that the machine does the real work and they’re just tidying up the output it generates when prompted, they naturally don’t merit high wages or indulgent perks, even if getting 30,000 lines of code regurgitated from the mashed-up contents of Github and Stack Overflow working is more cognitively tasking than writing that code from scratch would have been.)
It doesn’t matter what they claim if they simply can’t get the people to babysit the AI codebase or the AIs for less money than the original ones who didn’t have to deal with AIs and their output used to cost.
As a pretty senior dev who spent a lot of my career as a contractor mainly coming in to unfuck code-bases seriously fucked up by a couple of cycles under less experienced people, if I was pitched work to unfuck AI work I would demand a premium for my services purelly because of it being far more more fucked up in far harder to follow ways than the work done by less experience humans (who at least are consistent in the mistakes they make and follow a specific pattern in how they work) even without any moral considerations (on principle I would probably just not take a contract with a company that had used AI like that).
I mean, I can see their strategy work for junior devs, but that kind of reducing the power of specialized workers was already being done against junior devs using “outsourcing”.
My prediction is that it’s just the latest buzzword on the pile of buzzwords and by 2028 a new one will pop up and the only time you hear “AI” will be in the line of “Hey, remember when everyone was talking about AI?”
Before AI it was “The Cloud”. Before the cloud it was “Virtualization”. They’re saying all the same things about AI that they said about the cloud and virtualization…
I guess the real money is inventing the new buzzword that sales people can say will make your business faster, more agile, and more efficient. :)
I think it’s a real shame because all three of those things you mention are useful. The problem is that once they become a buzzword, then everything needs to be done using that buzzword.
Cloud has been misused to hell and back, and I have no doubt AI will too.
“AI-powered cloud software virtualization”
It’s just Kubernetes.
The Cloud is still a thing though. As is virtualization
And AI (LLMs, media generation, machine learning) are going to stay a thing as well.
Yeah, but nobody talks about them solving/causing all the problems. :)
Yeah, there’s generally a kernel of value wrapped up in all sorts of bullshit.
Some with the .com boom, obviously here we are with internet as a critical infrastructure, but 1999 ‘internet’ was a mess of overhype.
Before AI it was “The Cloud”. Before the cloud it was “Virtualization”. They’re saying all the same things about AI that they said about the cloud and virtualization…
So you’re saying AI will make a measurable (arguably net positive) impact and forever change the way we do things in our day to day, just becoming a standard toolset offered by many providers? Because I’d argue that’s what virtualization was, as well as the cloud to a lesser extent. Hell, I’d be hard pressed to be convinced on virtualization being a bad thing (not as much the cloud tho, that has some solid negative arguments).
If you’re trying to shit talk AI, you’d be better off comparing it to block chain than cloud/virtualization, since the latter two are an integral part of a large amount of the work we do, and the former is mainly for illicit drugs/activities and stealing money.
agreed, virtualization was one of the best things to happen in IT since the dawn of the internet. i can’t even imagine how much less efficient and reliable datacenters and the entire internet would be without it. Not at all comparable to AI.
i actually work for a company that does very little virtualization now and it’s fucking awful.
I think the comparison is apt, it’s not that LLM is useless, it’s just that, currently, it’s insanely overhyped. Just like the .com had irrational companies that evaporated but underlying tech was useful. Just like in-house servers were considered to be dead with everything being cloud hosted, now there’s recognition of a trade off. Just like there was pressure to ship everything as an ‘.ova’ and nowadays that’s not really done as much.
An appropriately used level of LLM might even be nice, but when it’s fuel for the grifters, it is going to be obnoxious.
LLMs may be overhyped, but the point is virtualization was not.
Virtualization as a ‘platform’ was a bit overhyped, hence my ‘.ova’ comment. There was a push for a lot of applications to exclusively ship as a whole virtual machine, to create OS variants dedicated to the purpose of running single applications. For a lot of applications it was supremely awkward, because app developers ended up having to ‘own’ things they didn’t want to own, like the customer network configuration.
Virtualization as a utility has of course persisted, but it’s much more rare for a vendor to declare their ‘runtime’ to be vmware than it once was. Virtualization existed in IBM for a long time, vmware made it broadly more available and flexible in the PC space, and then around mind 2000s things started to go a bit crazy with ‘virtualization is the runtime’.
Now mind you, compared to dot-com or ‘big data’ it was trivial, but it was all a bit silly for a time there.
Everytime I see this kind of hype pop up I think back to when there was this great announcement from Silicon Valley about a “revolution in transportation” and it turned out to be the Segway.
“People will be designing cities around this!”
It could have had an impact if it hadn’t been $5,000…
And, BTW, not even the coolest thing Kamen designed…
I think the next buzzword teed up is “quantum”
Looking forward to the infinitely scalable quantum AI blockchain cloud virtualization E2E P2P VPN micro-services!
Get this person a golden parachute, stat
🤢
Yeah, it’s already happening…
Same reason as forcing people back into the office even though it’s the solution to a number of serious issues affecting society:
Investors/banks have tons of money in these markets and are incentivizing/forces companies to adopt these policies to prop up the markets, whether it is in their interest or not.
Oh, yeah, we have that too… we want people in the office because collaboration! Synergy! etc. etc.
“How does that work if you want everyone using AI?”
“. . .”
Oooo hot take time: I’d rather work in an office again than be forced to use LLMs.
I’d rather use the AI than go back to the office. The AI doesn’t care if I’m wearing any pants.
I’d rather correct LLM hallucinations than some of the crap front line tech support tells people. :)
“The marketing and salespeople were enthused by the possibilities of working with these new tools, he added.”
[sigh] Because of course they were. Those people couldn’t find their own arses even if they used both hands.
Late stage capitalism rewards management for any appearance of change. It really doesn’t matter whether the results of that change are good or bad. And even a CEO who keeps destroying companies can always find a similar position elsewhere. The feedback loop is hopelessly broken.
Reminds me of the song Just Movement by Robert DeLong
Does he still have a company at all?
This type of shortsightedness should be punished. I mean AI can be useful for certain tasks but it’s still just a tool. It’s like these CEOs were just introduced to a screwdriver and he’s trying use it for everything.
“Look employees, you can use this new screwdriver thing to brush your teeth and wipe your ass. “
You can use this new screwdriver to fuck yourself. We’re working late boys!
