Its not that they aren’t intelligent its that they have been RL’d like crazy to not do that
Its rather like as humans we are RL’d like crazy to be grossed out if we view a picture of a handsome man and beautiful woman kissing (after we are told they are brother and sister) -
Ie we all have trained biases - that we are told to follow and trained on - human art is about subverting those expectations
Why should I assume that a failure that looks like a model just doing fairly simple pattern matching "this is dog, dogs don't have 5 legs, anything else is irrelevant" vs more sophisticated feature counting of a concrete instance of an entity is RL vs just a prediction failure due to training data not containing a 5-legged dog and an inability to go outside-of-distribution?
RL has been used extensively in other areas - such as coding - to improve model behavior on out-of-distribution stuff, so I'm somewhat skeptical of handwaving away a critique of a model's sophistication by saying here it's RL's fault that it isn't doing well out-of-distribution.
If we don't start from a position of anthropomorphizing the model into a "reasoning" entity (and instead have our prior be "it is a black box that has been extensively trained to try to mimic logical reasoning") then the result seems to be "here is a case where it can't mimic reasoning well", which seems like a very realistic conclusion.
I have the same problem, people are trying so badly to come up with reasoning for it when there's just nothing like that there. It was trained on it and it finds stuff it was trained to find, if you go out of the training it gets lost, we expect it to get lost.
That's apples to oranges; your link says they made it exaggerate features on purpose.
"The researchers feed a picture into the artificial neural network, asking it to recognise a feature of it, and modify the picture to emphasise the feature it recognises. That modified picture is then fed back into the network, which is again tasked to recognise features and emphasise them, and so on. Eventually, the feedback loop modifies the picture beyond all recognition."
After I got laid off in late 2023 I had a devil of a time finding work (despite having AI experience) to the point my unemployment ran out -
And I was 20 years as a dev and tech lead and full stack, (never had trouble finding work) including stints as a leading EM and CTO, I’ve been an industry award winning innovation lead,a digital studio director, switching tech stacks and cloud certs all my career..mentoring juniors and doing podcasts and writing white papers etc, but peanuts - nothing
Getting ghosted by 25 year olds in interviews and doing rounds and rounds and leetcode and all that but no success- for example I had a 7 round interview with NBCUniversal in 2023 and then got ghosted (I probably doged a bullet since they had subsequent layoffs)
a 12 month stint with nothing - we lost our savings as my wife got laid off too
Since then I pivoted to AI and Gen AI startups- joining incubators and finally got some work or at least cofounded some AI startups- now money is tight and I dont have health insurance but at least I have a job… it sucks as a over 45 year old as I have so much experience but no one cares.. still dont have much grey hair so I can pass for 40 to get noticed
No one stable is hiring or your resume just goes to a dead letter queue or is lost in ether or lost amid all the ai generated resumes out there - young ML and PHds and people under 40 seem to be getting work in Gen AI but thats about it
networking is the only game left and most good recruiters I know got laid off too
At least I’ve built up production experience in agents and context engineering RL, pytorch, langchain/LangGraph, RAG, KGa, etc python, BAML, LLM and LLMops to add to my years of full stack work
The number of just straight out NO's I've gotten has been surprising and disheartening. I've been a developer for 20 years and genuinely don't think I've gotten an outright "no" before this year. Usually just no response. I've gotten maybe eight or nine actual rejections this year. It's honestly worse, emotionally.
Ideally I want a job outside my wheelhouse to learn and move forward, but it seems like no one these days is interested in any sort of training.
There have been two gigs I was really excited about that seemed more-or-less exactly what I have spent the past decade doing and they've BOTH actually replied that they didn't think my skills were a good fit. I genuinely have zero idea what you're looking for if the literal perfect fit isn't it.
Before COVID I would regularly apply for jobs, almost always get them, and decline largely to keep my interview skills fresh.
This last year of looking has been the total opposite. I've applied for umpteen places, and gotten a little bit of email back and forth, and a single interview (it's in a couple days, wish me luck).
Idk about interviewing, but there are many benefits to opening fake job listing (gathering a database of people, keeping track of people looking for jobs, etc) which is why people do it. Data is valuable.
It is one trick that allows management to lie to everyone, and to implement downsizing. Fact is that A LOT of US businesses became unsustainable due to the tariffs (which Trump totally did for "US first" and not, I repeat NOT to get an extra tax started for him to spend), the retaliatory tariffs, followed by Trump doing the largest mass-firing in US history, presumably because this is listed in introductory economics textbooks in the chapter "what caused the great depression".
So a LOT of businesses are now in the position that they have to raise prices significantly into the worst market in decades. So they'll get significantly less revenue and they'll have to go into overdrive on saving money. That means no hiring, layoffs, price hikes, shrinkflation, ... the whole thing. They have to do a lot less with a lot less.
How does management respond to this? Well, management generally isn't competent. Their only job is to negotiate, and now that will really be put to the test. The smarter ones know that negotiation doesn't even matter under these circumstances (since it's a fixed pie being divided: someone has to lose). So they're maximizing their runway and getting out. And first, of course, they lie. They tell people even inside companies that they're hiring, even to the point of having interviews. They post job postings, because that's part of doing covert layoffs: they replace full time employees by temporary ones, even interns. They wait with price hikes until they're through inventory. They notice they've signed long-term contracts with Walmart that they cannot fulfill. And so on. So they lie to maintain their reputation for after the crisis (so on their next interview they can believably claim "I could have saved the company, but I found a better opportunity ...").
So I bet we'll be seeing a LOT of managers, especially higher up, suddenly decide they need to find a new job, and when it turns out that doesn't work, take a 2-3 year excuse to take a break.
As for replacing these people with AI: they're not spending ... and AI is expensive. Sure there's startups using AI, but larger companies are just firing people and not replacing them. Certainly they don't see the current period as a good time to change ... anything.
Governments should be "countering the crisis" and hire, according to economics textbooks, and increase social spending with the savings from the decade past ... except ... there are no savings from the decade past. So governments are firing, laying off, saving on healthcare, and so on and so forth. The US obviously has everyone's attention but the UK is doing the same (frankly, worse) and so is the EU.
Trump will be the most desperate manager of all, doing anything and everything he can to delay this from happening until the midterm elections in November, like lowering interest rates. But the thing everyone needs to remember about interest rates: they only lower for a good reason. The problem for Trump is ... either he delays this to beyond November or he becomes a "lame-duck" president, unable to do anything.
Two things: (1) Trump doesn't control the prime lending rate (and that's not the interest rate, either); and, (2) he's rapidly headed to lame duck status, already.
Unfortunately "lame duck" status doesn't mean nearly as much under today's radical Republican model of governance.
Namely, where the President does things unconstitutionally and illegally, like creating new taxes on Americans, refusing to execute laws saying money must be spent achieving a goal, ordering the persecution of opponents, etc.
Meanwhile, the new conservative majority of the Supreme Court refuses to act claiming impeachment is literally the only remedy, and a small minority of Republican legislators prevent impeachment from occurring.
Then I'm not posting job openings at all, and I'm working my personal network extra hard to get a pipeline of actual candidates setup.
And it's not just the ATS doing this -- oooodles of goobers who are nowhere near qualified are using AI and slamming the hiring pipelines. I'm already leaning on the personal networks because of that, anyway, but Parent Poster's idea just means I push for that explicitly instead of informally.
See also: layoffs.fyi -- the market is just SATURATED with quality talent
> ... And it's not just the ATS doing this -- oooodles of goobers who are nowhere near qualified are using AI and slamming the hiring pipelines. I'm already leaning on the personal networks because of that, ...
> In the graduate job market, the "goods" are job applications. When employers are flooded with thousands of AI-generated, indistinguishable cover letters, they lose the ability to identify the high-quality candidates who invested time and effort. The signal is drowned out by noise. Just as buyers in Akerlof's market stopped buying used cars, employers are stopping the open hiring process. They retreat to offline networks and nepotism to find people they feel they can trust. The issue is that if everyone sounds perfect on paper, the only signal left is a personal introduction.
That will never happen. We've got employers making employees sign ridiculous Non-compete, non-disclosure, non-disparagement, agreements. Do you really think this will ever happen? We've collectively decided that corporations are people and people have freedom of speech. We've spent years doing the H1B dance. (I'm not anti-immigration. I'd love to simply give them citizenship. I've met some amazing people on H1Bs. The system is exploitive to US workers AND the H1B recipients). Face it, as someone who is actually doing the work of IT, you lost a long time ago.
Amazon had a whole paper about how bringing migrants and diversity makes unionization less likely and successful.
It's part of the point.
Also it's not just the US. Here in Europe it's the same.
Hell the government spent it's money trough the employers org training Moroccans to become programmers to then bring em in.
I knew a fair few programmers who believe this is all just for low skill low paying jobs as if paying better wages for those is impossible and as if those deserve to be undercut.
The weird thing is how this kind of plainly rightwing economic rhetoric is often masquerading as left wing.
The only silver lining in this mess is that it does seem like this was finally the push for my industry to start seriously pushing for unions. Of course, corporate won't let it come easy but I only see momentum building now.
> Ideally I want a job outside my wheelhouse to learn and move forward, but it seems like no one these days is interested in any sort of training.
No one has the resources left to pay for it, that's the thing. Clients are cutting budgets because no one is buying their stuff, so everyone is looking for seniors and above only, and replacing juniors/intermediates with AI.
Trickle-down ideology is now beginning to eat itself, the ouroboros is complete - turns out, eventually the entire economy will crash down on itself when people can't afford things. Even Henry Ford already knew this...
... which was needed because no/less profit can and does send the stonk price tanking. That's also a "there are no resources left to do the right thing" - the stonk market dynamics are a part of what is eating our economies alive.
No, because software engineering interviews are a specific skill and you need to be on top of it - eg leetcode, sys design and soft skills and hiring managers are making everything harder
Ie its a race to the bottom - there is no gold standard for a productive software engineer who is transferable to another company that really people trust
How do you reckon? My understanding is that they're valid perm positions that legally must be offered to US citizens before they can be filled by anyone else.
it's really, really easy to say that applicants don't meet the requirements, since hiring is sooo subjective. "none of the applicants had 8 years and 2 months experience in this company working on this team, so they won't be able to perform the duties".
or, if they do have a flood of qualified applicants (the point of the website), they can simply not go forward with PERM at this time for their selected candidate. PERM roles aren't new jobs in the sense that the employer is "down" one employee if they don't fill it. PERM roles are simply a status change for an existing employee.
Nominally - although perhaps more often barking and not serious - is if you're a super duper qualified US citizen, you either can find a job there, or, you sue the company for discrimination, either way you get paid. The latter being the part that's highly hypothetical and potentially not realistic.
I work in a FAANG as a SWE. I'm not the on selecting candidates, but I do interview a lot of them. I'm pretty sure I've never seen a 50+ year old candidate. 99% of the candidates have less than 10 year experience, and are in their 20s or 30s.
So yes, ageism is real. I suppose the company gets away with it by saying they look for candidates with 3-10 years of experience?
At that stage, I think I would pass Leetcode with 2-3 months of practice, and I don't mind putting the work if this is what it takes. I'm just not sure I'd be given the chance.
I was a laid off in the beginning of the year from my remote job, landed several interviews, and I found a new job in <2 months. My resume is less impressive than yours, ~10 years of experience.
I was able to land interviews with some remote companies. I used to work at Shopify, a got some interviews at Ruby shops from that.
Some possibilities:
1. Ageism, this is a distinct possibility.
2. You held very senior positions. I think a lot of people don't like hiring people that were more senior than them. So that CTO is being held as a negative. They are not saying "Hey I get the experience of an EM and a CTO in a Senior Engineer for a bargain salary", they are worried you will overshadow them. This is sub optimal behavior for companies.
3. Talking to people I think non-tech markets look like they are doing fine. People I know in Rochester, Syracuse, and Cleveland aren't having issues getting jobs. I think the huge layoffs in big tech have left a big supply in tech cities to less demand.
> 2. You held very senior positions. I think a lot of people don't like hiring people that were more senior than them. So that CTO is being held as a negative. They are not saying "Hey I get the experience of an EM and a CTO in a Senior Engineer for a bargain salary", they are worried you will overshadow them. This is sub optimal behavior for companies.
---
Not only is there the overshadow worry, but there's the overqualified for the position.
You were CTO once, and now you're a IC again... are you looking for a CTO position still?
I had an experience (post dot com crash) where the team hired a senior engineer... who left the team within a year to be an engineering manager somewhere else in the company and our team was back to interviewing for the position again.
From our team's perspective we wasted the time interviewing and onboarding a person when they job-hopped (even within the company) in under a year. Despite being qualified as a senior engineer it wasn't what they wanted to do.
To that end, overly qualified candidates are similarly risky to hire as under qualified ones.
I've also been in situations where someone in a senior or management position is hired with previous Big Tech or startup experience and tries to make the regional retail company's internals into a Big Tech engineering department which ended poorly for everyone involved.
---
To your third point, the desire to stay on the coasts and around Big Tech companies is also a thing. Being willing to move to and within the midwestern states (some companies have return to office... others don't want to create a tax nexus in another state - if you don't hire anyone in California or Colorado you don't have to follow those laws... so hire everyone in one state and only deal with one state's payroll tax and insurance options).
> I had an experience (post dot com crash) where the team hired a senior engineer... who left the team within a year to be an engineering manager somewhere else in the company and our team was back to interviewing for the position again.
Same reason I struggled to get a job at McD's or Home Depot years ago when laid off. They knew I'd be gone ASAP and that I probably wouldn't eat shit the way the local rube demographic would. These are crappy service jobs but they still want you to stick it out for at least 6-12 months.
My career started in the south and then we moved to Seattle - it feels like everyone is laid off in Seattle - I could probably find something in the south again
can't comment on west coast but from perspective in NYC, i think the market for SWEs willing to work in office here is very good. might be different since there's lots of different industries hiring for SWE roles in addition to typical tech "startups" here.
i've noticed that folks who want to work remote having a tougher time if they're looking for tech jobs. makes sense if you look for jobs at a local non-tech company, you might have better luck.
generally seems like remote jobs have the most competition so if you can find jobs localized to your market, you will have more luck there.
4. Overqualified can also be seen as you are seeking this job purely temporarily. While continuing to seek better. You may be gone within weeks. A more alert project might seek to use your experience at low cost for that little while. But a harried manager already had a problem and might see it as now having two problems.
Being overqualified is indeed a real phenomenom. Employers can think your too expensive, or that your desperatem and that you might not stick around long. All valid concerns.
I don't mean this as a judgement; I am genuinely curious: with a career like that, how are you out of money after twelve months? Do you have unusual expenses? I am 31, currently making $180k (the most I've ever made) as a bog standard 'senior SWE', and I have nearly a million dollars. I am not frugal, I don't think; I just bought a $900 pair of sunglasses last week, and that wasn't a huge outlier expense. I have at least twenty years of savings right now. What is eating up your (and any onlookers', if you care to comment) budget?
It's also possible that OP means short term savings kept for circumstances like these, and long term investments bear more cost (e.g. keeping money in a money market fund vs in stocks has different tax implications when you take the money out).
Or good investment planning. I only recently built up savings after spending my entire career maxing out 401ks. If I was laid off, I'd only have like 6 months or so of savings before I run out despite having a coastFIRE/leanFIRE NW.
Yes, there are hardship withdraws from 401ks, but the older you get, the more retirement account means retirement account. Meaning, it becomes more clear that the money in that account needs to be left alone until things get dire. You're not going to be more employable in 20 years.
A 401k and an IRA together constitute only $31,500 this year, and less in each previous year, before taxes. That shouldn't be the main factor. FWIW I was including those accounts in my 'savings' number.
Getting replaced by H1Bs after a year and then being out of work for six months (because ageism), four times in a row--that'll do it! Especially now since H1Bs replacing you makes you look like a job-hopper so your job search becomes more impossible every time.
If botox is required to secure an engineering job when you have a long proven track record of solving hard technical problems successfully- the company is a fraud and the jobs are fake.
“Companies” don’t hire people, a few human individuals acting autonomously, with their own beliefs and biases, are ultimately responsible for bringing someone new onboard.
Once you hit 40s just get the botox anyway. Looks better. Only lasts 3 months or so. Good for a few rounds of interviews.
The way hiring works is a reflection of company culture. If they can’t execute a hiring process that hires the most skilled and competent instead of the youngest and best looking, guess what kind of coworkers you will have to deal with?
Anyways, I am in my 40s- I am an amateur strength athlete and eat healthy and already look almost inappropriately healthy and youthful for a sedentary job- enough so that I feel sometimes I’m not taken as seriously because I don’t look like a stereotypical elder nerd.
An ageist company culture does not mean the company is a fraud though or that the jobs are fake. It’s all very real, you are just rejected, that’s what makes it so frustrating, you could be living the good life if only those people weren’t so ageist. Now imagine being a POC and your whole life is like that, just racism everywhere keeping you out.
In both cases- ageism and racism, the employer culture is hiring less qualified workers based on criteria other than skill and competence. That necessarily means they have a lower caliber of expertise on average, as they are artificially decreasing the pool size of potential candidates. Moreover, if they have a discriminatory hiring practice in the first place, it is very unlikely that they are really selecting for competence even within the preferred demographic. For an extreme example, look at the current US administration, hiring only people with "Mar-a-lago face" combined with them needing to have a particular ideology and extreme lack of ethics, necessarily makes virtually all of the hires incompetent. They aren't, presumably specifically seeking incompetent hires, it's just the only people left given the fact that they select for things other than competence first.
I hire software engineers, and given all of the discrimination in the field, my willingness to hire women, minorities, and older people gives me a substantial advantage over those I am competing against that have discriminatory hiring practices. Although I'm not selecting for discriminated against groups, in practice I hire almost exclusively from them, because there are a surplus of extremely qualified people in those demographics that are getting rejected from everywhere else. I'm sure there are a ton of competent young white men in software engineering, but they're never among my top applicants, presumably because they've already been hired elsewhere.
This widespread discrimination pre-loads the applicant pool, such that the effect is enormous: you are getting substantially inferior employees if you have discriminatory hiring practices, because so many others do to, so your "preferred demographic" is already depleted of competent applicants.
Possibly besides the point but I wouldn't consider botox "taking care of yourself", taking care of yourself is staying in shape, eating healthy, and maybe dressing well and practicing good hygiene.
Can we just pause and consider what happened to this industry where this is the advice.
This is the natural consequence of marketing and fundraising completely destroying engineering. It's not about real things anymore. It's about image. It's an industry of carnival barkers turning everything it touches to shit.
The industry had turned to shit long ago when we deviated from the path of the UNIX philosophy and fell into cargo cult frontend programming. We had the most powerful UI in the world, the command line, and we gave it up for pretty pixels and shiny buttons. We have lost what it truly meant to be an engineer, now just look at the kinds of people entering the industry today: No passion for the science, just chasing cash and quick thrills. And you wonder why our interview process has gotten so shallow?
Do you have a thesis for why that might be happening? Ageism? Overqualification? The next generation of hiring managers not knowing what to do with you? Past experience being deemed irrelevant to modern SWE problems? Is it all just a bad market? Your profile strikes me as the last one that would struggle with landing a gig.
"Now it's different". Is it really? One aspect of these stories is that we all do age and we do climb into higher positions. So we run into "too old" and "too qualified" and "will leave asap for something better". And the reason is not that the economy or job market are different. The reason is that we are older. Nothing wrong with both happening at the same time but this "I never ran into this problem before" does have a common difference which is "I was never that old before".
I use "old" for shock value here while being totally in the camp that companies would do well to find ways to use older, more experienced talent at a lower cost than "normal". (Even after I have run into "older" as inflexible and a pain to work with.)
But there is also another problem in many of these reports: Older and still applying by sending resumes in response for job postings! If you are older you should have a network you can use. If they find nothing for you - or don't care to talk to you, then THAT is the better signal. And still not necessarily a signal about the job market.
Not sure what you are on about - I have a giant network in linkedin and it did help me find a job eventually…
You only cultivate a regional network if you live in the same spot for years - i moved from the south to Seattle - Im not moving back to the South- so my network is mostly useless
I have for example a network with OpenAI, Google and Anthropic - all it got me was getting to the team match stage with Google followed by denial after almost 6 months in the interview process
As for OpenAI and Anthropic you need to have success in the AI startup world+network+ be young (old is only allowed id you have massive career success eg VP at Google or Amazon)
Yeah, this ageism factor mixed with how unstable in general my industry is had me adjusting my long term plan to focus on being able to survive off my own products. I'm still far off from that point but it's probably best to plan my mid-late career before industry locks me out...
Doesn't help that I've tried to specialize for quite a while but simply can't stay around long enough without layoffs happening. Hard to be a domain expert when companies let you go ever 3 years. T shaped generalist it is, then.
What’s admirable is that you’ve actually applied to jobs below your pay grade. Most people won’t because doing that grind in your 40s is actually hard especially if you have kids.
My brother who is nearly 50 and has worked in tech since the dot com boom, got laid off in January and couldn't find a job until last week. This job, too, was just a contractual position at his old Fortune 50 firm.
He has an engineering degree from one of the top 5 engineering colleges in India, a Master's from one of the top 5 engineering schools in the US. He built some of the systems that form the foundation of the entire call center industry.
And now he pivoted to GenAI and has dozens of very impressive public projects including some heavily starred open source repos
The dead comment replying to this said "Start your own company," which I hear a lot. The problem is, the #1 thing YC looks for in a founder is they're wealthy enough to absorb risk. You don't take on massive risk from a position of financial insecurity.
Are you ready for great age of suffering? Because there will be no saving, no UBI, no plan B. This is it. This is the end of careers and beginning of centuries of misery. AI will replace everyone except around 3%. If you aren't one of those 3%, start accepting the reality of infinite pain.
Sorry for them- after I got laid off in 2023 I had a devil of a time finding work to the point my unemployment ran out - 20 years as a dev and tech lead and full stack, including stints as a EM and CTO
Since then I pivoted to AI and Gen AI startups- money is tight and I dont have health insurance but at least I have a job…
> 20 years as a dev and tech lead and full stack, including stints as a EM and CTO
> Since then I pivoted to AI and Gen AI startups- money is tight and I dont have health insurance but at least I have a job…
I hope this doesn't come across as rude, but why? My understanding is American tech pays very well, especially on the executive level. I understand for some odd reason your country is against public healthcare, but surely a year of big tech money is enough to pay for decades of private health insurance?
Not parent commenter, but in the US when someone’s employment doesn’t include health insurance it’s commonly because they’re operating as a contractor for that company.
Generally you’re right, though. Working in tech, especially AI companies, would be expected to provide ample money for buying health insurance on your own. I know some people who choose not to buy their own and prefer to self-pay and hope they never need anything serious, which is obviously a risk.
A side note: The US actually does have public health care but eligibility is limited. Over one quarter of US people are on Medicaid and another 20% are on Medicare (program for older people). Private self-pay insurance is also subsidized on a sliding scale based on your income, with subsidies phasing out around $120K annual income for a family of four.
It’s not equivalent to universal public health care but it’s also different than what a lot of people (Americans included) have come to think.
As CTO I wasnt in a big tech company, it was a 50 person digital studio in the south
my salary as was 275k at the highest point in my career- so I never made FAANG money
Yeah. It is much harder now than it used to be. I know a couple of people who came from the US ~15 to 10 years ago and they had it easy. It was still a nightmare with banks that don’t want to deal with US citizens, though.
As Americans, getting a long-term visa or residency card is not too hard, provided you have a good job. It’s getting the job that’s become more difficult. For other nationalities, it can range from very easy to very hard.
If you don't have a university degree, most of EU/EEA immigration policy wants nothing to do with you, even if you're American or have several YoE. Source: am a self-taught US dev who has repeatedly looked into immigration to northern/western Europe over the years. If anything it continually gets more stringent every time I look. Forget looking for jobs, there's not even visa paths for most countries.
But isn't the same true for the US? To me it seems it's pretty similar both for Europeans moving to the US and Americans moving to the EU: have higher education, find a job, get a work visa...?
No clue, I don't know much about US immigration policy other than that it is, by all accounts I've heard, a nightmare. I know (from past experience) that Canada has no-degree-friendly paths, but I have no reason to expect the US (with its current set of policies especially) would be that progressive/open.
Yeah it depends on which countries you're interested in. Netherlands, Ireland, and the Scandinavian ones are on the easier side as they don't require language fluency to get (dev) jobs, and their languages aren't too hard to learn either.
Strictly speaking, Finland is nordic, not scandinavian. And their language is entirely different. I was under the impression Fins and Estonians are happy when people try to learn. Am I wrong?
I made a career out of understanding this. In Germany it’s quite feasible. The only challenge is finding affordable housing, just like elsewhere. The other challenge is the speed of the process, but some cities are getting better, including Berlin. Language is a bigger issue in the current job market though.
Counter: come to Taiwan! Anyone with a semi active GitHub can get a Gold Cars Visa. 6 months in you're eligible for national health insurance (about 30$ usd/month). Cost of living is extremely low here.
However salaries are atrocious and local jobs aren't really available to non mandarin speakers. But if you're looking to kick off your remote consulting career or bootstrap some product you wanna build, there's not really anywhere on earth that combines the quality of life with the cost of living like Taiwan does.
Taking a 75% pay cut for free Healthcare that costs 1k a month anyway doesn't math. Not to mention the higher taxes for this privilege. European senior developers routinely get paid less than US junior developers.
I'm sorry I won't share much details, I don't think much is public on Vsora architecture and don't want to breach any NDA...
From their web page Euclyd is a "many small cores" accelerator. Doing good compilation toolchains for these to get efficient results is a hard problem, see many comments on compilers for AI in this thread.
Vsora approach is much more macroscopic, and differentiated. By this I mean I don't know anything quite like it. No sea of small cores, but several more beefy units. They're programmable, but don't look like a CPU: the HW/SW interface is at a higher level. A very hand-wavy analogy with storage would be block devices vs object storage, maybe. I'm sure more details will surface when real HW arrive.
your use-case is NOT Everyones use-case..(working in depth across one codebase or api but instead sampling dozens of abilities across the web or with other systems) thats the thing
how is that going to work with my use case, do a web search, do a local api call, do a graphql search, do an integration with slack, do a message etc..
Does it matter ? if it's well defined, each of those would be a node in the graph, or can you elaborate ?
Dozens seems not that much, for a graph where a higher level node would be slack, and the agent only loads further if it needs anything related with slack.
Or I'm not understanding.
I tell people this over and over and over: every time you use a third party dependency, especially an ongoing one, you should consider that you are adding the developers to your team and importing their prior decisions and their biases. You add them to your circle of trust.
You can't just scale out a team without assessing who you are adding to it: what is their reputation? where did they learn?
It's not quite the same questions when picking a library but it is the same process. Who wrote it? What else did they write? Does the code look like we could manage it if the developer quits, etc.
Nobody's saying you shouldn't use third party dependency. But nobody benefits if we pretend that adding a dependency isn't a lot like adding a person.
So yeah, if you need all of posthog without adding posthog's team to yours, you're going to have to write it yourself.
> I tell people this over and over and over: every time you use a third party dependency, especially an ongoing one, you should consider that you are adding the developers to your team and importing their prior decisions and their biases. You add them to your circle of trust.
Yes. If your shop is serious about security, it is in no way unreasonable to be building out tools like that in-house, or else paying a real vendor with real security practices for their product. If you're an independent developer, the entirety of Posthog is overkill, and you can instead write the specific features you need yourself.
We had created a sort of Posthog, but for product security analytics (1), and after 4 years of development I can confirm it's not something that you can easily create in-house.
If they have a HTTP API using standard authentication methods it's not that difficult to create a simple wrapper. Granted a bit more work if you want to do things like input/output validation too, but there's a trade-off between ownership there and avoiding these kinds of supply-chain attacks.
If you aim for 100% coverage of the API you're integrating with, sure. But for most applications you're going to only be touching a small surface area, so you can validate paths you know you'll hit. Most of the time you probably don't need 100% parity, you need Just Enough for your use-case.
To my understanding, there's less surface area for problems if I have a wrapper over the one or two endpoints some API provides, which I've written and maintain myself, over importing some library that wraps all 100 endpoints the API provides, but which is too large for me to fully audit.
Its rather like as humans we are RL’d like crazy to be grossed out if we view a picture of a handsome man and beautiful woman kissing (after we are told they are brother and sister) -
Ie we all have trained biases - that we are told to follow and trained on - human art is about subverting those expectations
reply