n_ary

Will actually open more ad opportunities of marketing more private contents and detecting preferences at higher level. This data will also of course cater to more advertising money from questionable industries, which in the end goes to Facebook coffer and they can shrug saying, they only provide a platform, whatever people are buying or clicking is whatever people want, otherwise they wouldn't be clicking those ads.

While this is most alarming and sad state of affairs for LLMs being peddled to lowest denominator of anything valuable to replace lonely-heart scammers with more platform engagement. I believe, this should pull away people from other scams which may have higher consequences(financial loss is major one).

Other than that, I wanted to say kudos to your very SEO combination of particular word associated to Facebook all over your comment. I am optimistic that, if this thread gets adequately ranked higher in searches, I want that combo of yours to come up the most.

> "One example I heard: Microsoft integrated AI into Word and other Office tools--and the first thing most users tried to do was figure out how to turn it off. That tells you something. These companies aren’t necessarily in touch with what people want or need."

I mean, the people who spend most of their time in MS Word and other office tool are the majority group who are not incentivized to do resume-driven-development, and less tool they have to master, less overall mental stress they need to bear. This group is interestingly also very difficult to train unless they are forced to, as this group is often protected by job security due to being government worker, regulatory duty(banks/insurance/audits/finance/medical etc.), union/work-council member, basic necessity, or employer lacks any incentive to invest in automation because re-training a biological being is much less expensive and more fail proof in the long term.

Microsoft is aware that, given the opportunity, they can replace a vast majority of these group in a heart-beat, but manipulating the incentive is difficult as of now. However, they are prepared and will continue collecting data(by adding these features and having data to show how someone was doing basic tasks and how someone taking the time to learn has delegated their duty to these tools) to be ready once they can deploy the automation.

AI will survive us. In fact, I am worried about the current trend and see feeding the mighty AI as being a mandatory thing in future. Unchecked monopoly has a way of mandating things that are net bad for everyone except the middle man. Previously, job search was nice, but now 100% of the job application require your linkedin profile url, if not provided, you get auto-rejection.

People with decades of experience in the trenches who recently got laid off(business failure, corporate greed cutting costs, restructuring ..) now are asked everywhere to submit a link to their github(no one knows gitlab/codeberg/sourcehut etc) full of portfolio projects! I talked to few academic friends, who are worried that their research work is now reproduced verbatim by two specific LLM HN really loves!

Unless, LLMs go the way of ads to survive and rely on SEO spam to retrain, a monopolistic capture will happen mandating that all useful content must be fed into common hubs where AI can happily ingest it but cumulatively no human expert will be able to use it(we all know the abysmal state of info retrieval) and LLMs as these become more popular will become ever so unreachable for common folks without lots of riches. For medium term, I see a Netflix/Amazon Prime Video play, LLMs as these get more popular(same way people mindlessly scroll yet lecture others of its harm), will raise prices and lock out people from the common good and serve specific beneficiary group(shareholder).

You should try with lowered temperature(0.1 or even 0.0). Usually accuracy improves dramatically.

Tbh, your job is secured as long as your boss is not brainwashed and FOMO by the infinite cash burning and lobbying to sell LLM as the all solving holy grail. Depending on how long it takes to brainwash all leadership into thinking, LLMs can do your job(despite reality, I mean it takes some time to fail and then additional to deal with sunk cost fallacy). One of my colleague went to work for a startup that promises to replace ALL general practitioner/family doctor within next 2-3 years by LLMs. While I hope they badly burn and go bankrupt, I am also worried about the future as this particular startup is also getting major backing from certain big deep pockets.

Gaming was sort of their major thing until the cryptocurrency wave gave them a major boost and GPUs were rare luxuries suddenly. But as crypto was fading and GPUs were suddenly becoming affordable, GenAI wave hit and again gave a major boost. Am curious how they will react to quantum(sadly no magical enterprise application apparent yet so the tail wind is not strong).

The number of times real human powered support caused me massive headache and sometimes financial damage and the number of times my lawyer fixed those because me trying to explain why they were wrong… I am not surprised that AI will do the same as the creation is the image of the creator and all that.

Are we sure that the damage is in “millions”? Could it not be just the same old “vocal and loud minority”, who will wake up tomorrow and re-activate their subscription before the lunch break once Cursor team comes and writes a heartfelt apology post (well generated by AI with a system prompt of “you are a world class PR campaign manager …”)?

Having some experience with weaker models, you need at least 1.5B-3B to see proper prompt adherence and less hallucinations and better memory.

Also models have subtle differences, for example, I found Qwen2.5:0.5B to be more obedient(prompt respecting) and smart, compared to LLama3.2:1B. Gemma3:1B seems to be more efficient but despite heavy prompting, tends to be verbose and fails at formatted response by injecting some odd emoji or remark before/after the desired output.

In summary, Qwen2.5:1.5B and LLama3.2:3B were the weakest model which were more useful and also includes tools support(Gemma does not understand tools yet).

Please share more about this half-cooked agent on Linkedin. I am getting very curious.

Correct. My dusty Intel Nuc is able to run a decent 3B model(thanks to ollama) with fans spinning but does not affect any other running applications. It ks very useful for local hobby projects. Visible lags and freezes begin if I start a 5B+ model locally.

> Is it just hard to maintain that level of enthusiasm over time? Is the sum of possible money just too desirable?

I do not recall the book/essay or the original author, but I recall the quote that “everybody has a price they can’t resist, find out that price…”.

If your enthusiasm product is successful and has potential to be milked, someone somewhere will figure out your price and eventually buy you out.

Here is a different take. The GenAI is just another revenue stream for big players with scale or resources.

At work, I was tasked with building some magical agentic stuff, working on it for a while, I realized that, HN shouts, oAI/xAI/Google/Amazon/Anthropic have no moats, there are oss models available. The actual fact is, the moat is the access to scale of resources(GPUs, power infra, network), which is very difficult to build for random joe's startup.

You must always rent a model access from one of these players(even OpenRouter delegates), and that is the exact moat.

GenAI solves problems of having more generalized solutions, so instead of a super customized secret sauce solutions for your product as competitive edge, you now build magic prompts for GenAI to take the same input and hopefully with enough GenAI ingesting the same data and coming to consensus, you get a reasonably useful output that previously your custom solution was capable of. Since you no longer have a custom solution, you now pay off the GenAI operators(the real moat of GenAI for all operators hosting it). In the process,you also sacrifice your competitive edge of that super secret IP and relying heavily on GenAI prompting correctly and numerous verification in each step with enough automation, which of course again costs money.

GenAI is the new hammer of visionary leadership and executives (a hefty amount of money has been burned to campaign and PR to convince these people) to use it everywhere, so the operators can ensure that they make some profit from the amount of money they sank on it. If you super impose the "AI" of current year to "Apps" in 201x era, where everything must have an "app", you'll suddenly realize that we've seen the same before and of course most apps need cloud... and as clouds have costs, apps became more subscription model instead of previous 200x era buy once, use forever.

Off-topic:

In 2024-2026

s/app/AI/

How the cycle repeats, just the deliverable changes.

I still use search engine. LLMs are great but often get outdated by 6-12 months. Usually, I search for random coding topics and asking LLMs will often reproduce the outdated top(most voted) answer verbatim. However, there are some nice folks at SO/SE who come back later and update the answers or submit a new one, but LLMs often don’t return these but continue producing various spin off of the top answer by modifying variable names or adding/removing comments.

Hence, search still remains my hope until SO and the likes decay.

Additionally, many search engines now already generate quick summaries or result snippets without a lot of prompt-fu, hence LLMs have actually become 40:60(llm:search) ratio day to day.

<quote:esperent> In theory, vibe coding can let someone with very limited technical expertise build complete apps, so understandably a lot of people are excited by it. </quote:esperent>

Actually it is booming. In bsky, X and Linkedin, I see another recipe/todo/budget management/profit tracking/SaaS starter template/landing page/people-to-follow-directory etc. being pumped out every single day. Before GenAI, this would be more like one partial feature per user every month, now post GenAI, entire product in weeks or even hours.

I believe, the indie entrepreneurs are making the maximum bang for the bucks with AI codegen compared to any other groups.

To correctly vibe code something useful, I find that I need to religiously give faith to the tool and forget all about SWE principles and best practices and instead treat it like a child who makes mistakes and corrections while the adult must not intervene or admonish too much but rather nudge it to right direction.

Also vibe code has a parallel feature, while the code is generating, you are also doing live review and correcting it towards right direction, so depending on your experience, the end product can be a bad mess or wonderful piece of creation and maintenance dream.

The issue with seasoned SWE is that, the moment a mistake(or bad pattern) is made, the baby is thrown with bath water.

For a tiered app like the one presented, 35k LOC is not really that impressive if you think about it. A generic react based front end will easily need a large number of LOC due to modular principle of components, various amounts of hooks and tests(nearly makes us 25-40% of LOC). A business layer will also have many layers of abstractions and numerous impl. to move data between layers.

The vibe code shines, when you let it build one block at a time, limit the scope well and focus. Also, 2-3 weeks is a lot of time to write 35k LOC. at start of any new project, LOC generation rate is very high. But in maintenance phase it significantly falls as smaller changes are more common.

So, “previously a shoot would cost 5-figure […] now a kid can do it for $500”.

“Soon we will no longer need musicians, due to rise of AI”.

“If a director can do it with AI writing I want X, Y, Z”

A strange pattern. In all three cases, are we getting a 1:1 quality or we are so lost in age of cheap quality that those does not matter?

I for one would not buy anything or pay a subscription fee, where the photo or the music was AI generated. If the creative content I want does not have a humane touch, it is not worth to me.

Also, while we are bemoaning the loss of fields and jobs, behind the veil lurks the age of wrong incentives. In previous era, when my parents were pursuing a career, a record breaking profit meant more expansion, more investment, more debt and new pursuit of possibilities. Now we live in a moral bankrupt feudal age, where despite record profits, layoffs and merger+shutdown are performed like a crazy to maximize stock price and top level bonuses. It all is just regression of society’s feudal lords too deep in greed to care about the future.

No one thinks anymore, here is my product, it is so cool and ground breaking, we must layoff more employees to also chase that another fad which will likely fail in a year or two but I get good bonuses and move onto another place before the consequences come due but then more layoffs can be done to pay for my mistakes. But eventually if enough people are being laid off, who will buy my products? How will my target audience pay for this thing if they don’t have a job? The economy does not work if the gains are not cycled again back to society and common folks.

Also welcome to the age of media control and mass manufacturing consent(thanks to Noam Chomsky for the concept), when enough media words and campaign can be bought to tell the feudal overlords what to do and what is the new trend in societal destruction in order to maximize their billion bonus packages in short term and leave the consequences to the future.

/rant off

While I sometimes hangout with non-Engineering folks, I see following activities:

- preparing various letters, reports, audits, notices, terms/contracts etc. by modifying or using very specific templates and Office tools are very well entrenched for fixed templates

- Financial reports, analysis on excel(yes, most professionals are trained and feel comfortable there, same as how we feel good in terminal/console), building specific forms with constraints and checks, data sharing, special report of field data collection and audits, technical specifications/spec-sheets, quick complex calculations, think MS Excel as best candidate which everyone and there pet dog learned since elementary school

- Presentation, brochures, special fliers, quick reports, course reference/speech deck, lecture slides, product marketing/sales docs are mostly being handled by PowerPoint

Of course, we have numerous best-alternatives which can all replace these or “why not google docs”, but these people are trained on these specific ms office tools since dawn of civilization when people stopped cave painting and stone tablet based accounting in favor of computers. While we tech folks love to pickup and learn new tools, the other 99.99999% people see diminishing return adapting new tools. And lets be real, who wants to mess with the official template in some new tool that purchasing and procurement will not ever care about and scold the applicant back into the MS Office tools that already comes pre installed and licensed from their IT vendor?

But if those aspiring juniors stop asking those questions and the seniors not answer them in open web, neither LLMs nor juniors have new training data and both become stagnated. How do we solve this?

At my early days, I learned more reading code from much senior engineers and began to appreciate it. An effective seasoned senior writes a beautiful poetry that conveys deeper meaning and solves the entertainment(erm… business requirement) purpose. If the seniors retire and the juniors are no longer hired, then where do LLMs get new data from?

In all sense of things, I suspect we’ll see more juniors getting hired in coming years and few seniors present to guide them, same as how we previously had few db specialists and architects giving out the outline and the followers made those into actual products.

The article appears to be a rant and panic piece…

At this point, I am totally confused. When I attend expensive courses from Google or Amazon, the idea in the courses are that, tech has become sooo complex(I agree, look at the number of ways you can achieve something using aws infinite number of services), we need some code assistants which can quickly remind us of that one syntax or fill out the 10000th time of writing same boilerplate or quickly suggest a new library functions that would take several google searches and wading through bad documentations on another 5h of SEO spam or another 50 StackOverflow with same issue closed as not focused enough/duplicate/opinionated.

It is like, they want to sell you this new shiny tool. If anyone here remembers the early days of Jetbrains IDEs, the fans would whirl and IDE would freeze in middle of intellisense suggestion, but now those are buttery smooth and I actually feel sad when unable to access them.

Now, on the outside in news, media, blogs and what not, the marketing piece is being boosted a 1000x with all panic and horror, because a certain greedy people found that, only way to dissuade brilliant people from the field and not bootstrap next disrupters by signaling that they themselves will be obsolete.

Come to think of it, it is cheap now. First idea was to hire them when investments were cheap and disruption risk was high, then came the extinction of ZIRP when it was safe to stop hoarding them as no investment means less risk of disrupters, but if some dared, acquire and kill in the crib. Then came bad economy, so now it is easier to lay them off and smear their reputation so they can’t get the time of the day from deep pockets. Final effort is to threat the field by fake marketing and media campaign of them being replaced.

This panic drama needs to stop. First we had SysAdmins maintaining on-prem hardware and infra. But Aws/Gcp/Azure/Oracle came along to replace them to only move them up the chain and now we need dedicated IAM specialist, certified AWS architects, Certified Azure Cloud Consultants and what not.

Sorry for the incoherent rant, but these panic and “f*k you entitled avocado toast eating school dropout losers, now you’ll be so screwed” envy social media posts are so insane and gets so much weird, I am just baffled by the whole thing.

I don’t know what to believe, big tech telling me in their pretty courses and talks about how my productivity and code quality will be now improved, or the media and influencers telling me we are going to be so obsolete (and avocado toast eating dropout, which I am not). Only time will tell.

In the meantime, the more I see demos of impressive LLM building entire site from everyone and their pet hamster, the more number of Frontend engineering jobs popup daily on my inbox(I keep job alerts to watch market trends and dabble in topics that might interest me).

Off-topic: Mistral already had web search enabled for a while on there free tier.

Caveat: Mistral reasoning model on free tier is super slow(2-5 token/sec).

What do you mean 08:00PM. We have enough sunlight up into 09:47PM(21:47) during mid to late summer and darkness falls after 10:30PM(22:30).

Winter is… Sunrise at 08:12AM and Sunset by 04:00PM(16:00).

You can technically use this info to guesstimate my location ;)

I believe, you are a specific minority actually. What the OP described “gal flipping between FB/Insta” is actual reality around me when I am outside and look around me. Young and adult, all are flipping, the younger they are more TikTok/Insta, older they are more FB/X.

Even some of my family members who preached the bad of smartphones and internet are addicted to Facebook/Youtube and mindlessly scroll away while I observe the insanity.

It is sad and unimaginable for those of us who can resist the claws of the addiction, but for the rest 99%, it is just escaping the boredom/loneliness/social disconnect.

Social media wanted to connect people to genuine interactions over long distances, instead it has created a whirlwind of addiction media supply. Most people’s feed are very insignificant portion of their real connections and dominant following of content farms, celebrity gossips, meme, media misinformation, polarisation and of course 66% of ads trying to sell them something related to the most common trend of their feed at that hour.

I meant red(dead). Also losing a lot of valuation after jacking up to that point is not fun.

Also a business paying $1M/year in cloud bill is neither poor nor tiny. We have very different definition of tiny and poor.

Meta is very well suited in this AI race and will prevail well. Others will cut some loss and throw some ashes sometime soon.

The real answer is, leaders chase trends, so to keep them satisfied you need to also chase trends.

Some advises can be:

* learn prompt engineering to impress your boss and next employer

* adapt AI IDE at work, of course don’t go for cline or freemium ones, go for max expensive tier of Cursor/Windsurf/v0 etc

* Take some expensive(more the better) courses and workshops on Agentic topics, build some small projects. Justify your expenditure by citing being more well prepared for AI transition and adapting to new paradigm to strategically beat the competition businesses who will be extinct without employees with such trainings

* build some proof of concept projects to convert the smaller trivial projects to use agentic workflow. Then show these to your boss and put these up on github for future reference

* learn to train smaller models on your own internal documents, build a chat interface on top and give access to your boss(trust me, they will be blown away and will sell this to their superior)

* seed fear in your colleagues’ minds by using AI stuff where possible

Think of this AI as a new trend we must adapt to, just like FE moved from jQuery to React. Life and work goes on, just this wave nudges the complacent bunch to finally get out of their comfort zone and learn something new or get left out.

> Steer away from skills like web development that are clearly getting eaten by LLMs

On the contrary, here in my corner of EU nearly 60% of new jobs are frontend or full stack.

Anything else left is mostly SAP consultant or DevOps.

I think, the whole WebDev is dead end is just false lies to dissuade new entrants. Literally most successful business with a digital solution is web stuff with some automation that would otherwise be ms excel sheets shared via email.

Also this whole panic over LLM is overblown, I know of some brilliant experienced people in other professions like Electronics, Mechatronics, Aerospace, Material Science and literally all of them are finding the job market “very difficult at the moment”. It is the bad global mood in general used deceptively by opportunists to spread false fears of their LLM/AI.

At the end of the day, an insurance seller has hundreds of concerning reasons to convince you why everything is dangerous around you and you really need their product. Now apply that to AI sellers.

Actually, all stocks are dead and tech heavy Nasdaq has entered the correction zone this year. It will be a real while before the stocks return to their pre-correction value. The AI hype drivers have lost massive value, hence the desperation will get immense, expect the “AI ready to replace expensive SWE/Lawyer/Entertainers/Creatives” narrative to start any moment now! Of course everything is smoke&mirrors so those will be charged at such premium that the naysayers will not be able to afford it and irrational leaders will buy it and cite immense success to maximise their C level bonuses and will early retire before the tower comes falling down.

I believe, we are beyond the point of “bad software written, bad software deployed, business as usual” point long ago when AWS/GCP/Azure became an important requirement in job description.

A bad piece of software can be decently hidden by burning more money in cloud bills, which gives the inflated sense to the leadership that their products are doing global scale ground breaking.

With AI, I would not be surprised if the quality actually improves and the cost comes down(or stays same). Of course, more bad software will be written by now many aspiring entrepreneurs to realize their dream idea of spotify clone, then sacrificing their life saving on complex cloud bills and ever so profitable rise of revenue of all cloud services citing this as benefit of AI while doing some more layoffs to jack up the stock prices.

The real revelations will come(it always does, nature and economy works in cycles), when excessive layoff caused damage will come due and now everyone will scramble to rehire people in few years. Unlike the Ford innovation of replacing horse carts, software is more prevalent in our every aspects of life, same as doctors and lawyers and civil service, hence we need to honestly play the game until the wave turns and then cash in by making in 200x killing just like the businesses are cashing in on right now.