The AI Fear is real

The AI Fear is real.

A lot of intelligent people discount AI, LLMs, and the massive progress with comparisons to statistics and other known but trivial methods…

…that they don’t always seem to be familiar with either.

Those same people also seem unfamiliar with what even the consumer-grade AI ChatGPT can do - let alone what’s “cooking in the labs”.

Some people have romantic explanations of how an AI cannot be creative, cannot understand complex scenarios, and not be able to comprehend “the bigger picture” like an “IT Architect.”

Folks, we’ve seen nothing yet.

What WOWs us with GPT-4 is an AI in its baby shoes.

AI just got started and is touching down on some people in the mainstream.

Most people are still clueless. Most people are not working with AI daily.
What is their competence to write about it, anyways?

Regurgitating some press releases and stories they heard “AI Influencers” talk about - that’s about it.

Churnalists spend 10 minutes with a dumb 1-liner prompt in ChatGPT and

call out, “ChatGPT sucks”.

when in fact, their prompt sucks.

That’s all, pure incompetence.

I don’t recall ever so many non-experts talking about a subject matter than in AI.

I guess that’s because the “required skill” is just English - not some programming language or technical topic that would require some “hard skills”.

I’ve only been exposed to AI prompting for about a year, so I can call myself a practitioner at least, not an expert.

But I disagree with a lot of the commentary online from people making their money in ways that AI will VERY soon replace.

This includes “fancy” jobs like software development, project management, content marketing and of course churnalism.

Those simple “well paid” white collar jobs, including artists or any kind.

Of course AI is being talked down.

What shall not be cannot be.

It’s too scary for some, incomprehensible to most. The revelation that not only general coding, but even advanced skills in software engineering will be replaced very soon is not an easy swallow.

I can imagine GPT-4 becoming even better than the 90% of code generation I use it for now - with decent prompts.

And a bit more magic sauce that my team and I am working on.

There’s no way back,
and
there’s no "but AI will never…"

We’ve seen nothing yet.

Embrace the change!

6 Likes

Continuing the discussion from The AI Fear is real:

what makes all this a real adventure is that we are in unchartered territories, often marching on egg shells, trying not to panic while learning and testing like crazy. No time in fact to listen to those scared and those of us applying as they time permit to get better in grasping the full potential of a life deeply enhanced through AI.
I can’t even test 10% of the tools but I extensively used chatgpt, built 2 python scripts calling GPT APIs without knowing to code (eternal coding beginner and building a bootcamp for the rest of us), wrote 500 articles often SEO not that friendly, translated my novel into english (200 pages), translated a great public domain book on Psychology in French (300 pages) but I was in a frenzy of unfinished businesses killing spree.
What made me so appeased and scared at the same time ( it’s an undefinable feeling, a mix of awe and chill) was the fact that I took care of difficult thing, of “impossible before stuff”, of “I need an army of freelancers to do this and that”.
It’s a new era, a start of a new civilization, and it’s great that so few become aware of it, we’ll still keep an edge before things get out of control once AGI is reached.
So, frankly, I don’t take much time listening to the people you mentioned although I understand it can make people cringe. But as I mentioned “The Power of Killing My Unfinished Businesses” in all sort of creative endeavour I started over the years would have not be possible. And this is where for me there is, as you mentioned, no turning back.

2 Likes

Is it possible that the government has not yet established a complete “unemployment insurance system” that causes people to fear that AI will take away their jobs? It is difficult for someone to embrace dedication amidst the fear of unemployment, so you should now go to the White House official website and ask: "When will the unemployment insurance benefit in the AI era be implemented? How much is the unemployment insurance benefit? How will you distribute the value created by AI ?”, personally think that Sam Altman’s “World Coin” is probably a good idea, it depends on whether those politicians can figure this out. :partying_face:

1 Like

AFAIK there is no helpful unemployment insurance in the US.

In Europe things developed so that a certain chunk of the population is even unwilling to look for work, being “OK” with the (quite substantial) payments you can get from unemployment, not for a great life style, but certainly survive.

Lots of talk about increasing that even, some suggesting universal basic income even.

There are parts of this I very much agree with, but also parts that are commonly misunderstood, especially the “unwilling to look for work” part. I’ve certainly had a couple of times of unemployment in my life and not one of those times were in any way comfortable. Obviously this is a UK perspective, but that’s at least mostly in line with EU regulations, of course.

If you are single, with no dependents, sure, it is easy to get a job that pays better than being unemployed, provided you don’t have a high rent/mortgage (in the UK, rent/mortgage is covered provided it is ‘reasonable’, like not renting a 3 bedroom house when single or a childless couple) before they calculate what you get paid to cover everything else. That thing of the rent or mortgage being a separately covered thing can be a big deal.

If you have kids or other dependents though, well, today’s Europe just isn’t really geared towards the days of yore where there was a sole breadwinner and a stay-at-home-mom. Wages are set at levels that really need both partners to be working just to afford a pretty basic standard of living, and thus it is very difficult to find a job that would pay enough to even cover the bills as a single worker, but with non-working dependents. Sure, government programs exist (through absolute necessity) to “top up” such wages to be at least equal to the unemployment payments, but these rarely cover the additional expenses of working in real terms such as travel costs, meals at work separate to the meals for the family at home, work clothing where it is not provided as a uniform, etc.

So often I have known friends and acquaintances that were actually worse off working than being unemployed due to the additional expenses (and notice I didn’t even include daycare if the dependents were kids too young to be left alone), and they still chose to work just to avoid being bored, or to feel useful, or avoid social judgements, etc.

This is not because welfare systems pay too much, but because far too many company wage levels are exploitative, aimed at school leavers with few bills, and if you are not a school leaver with few bills, they’ll happily replace you with someone that is if any are available and able to do the job. Wages are too low, including statutory minimum wages that have led to massive reliance on charities, food banks, etc.

However, I have a genuine concern that we will all learn a lot more about the welfare system in the coming decade, because there’s a genuine and significant risk it will crash. Welfare systems all tend to run into trouble when more than around 10% of the population need them. They are all based on 90% of the population paying the taxes to cover all the costs.

The AI Revolution (and I don’t mean revolution like a robot uprising, but rather in the sense of the Agricultural Revolution, or the Industrial Revolution) is going to create a massive shift in economies and economics. This is obvious.

The agricultural revolution was all down to simply learning how to do crop rotation, at which point mechanization of farming with better machines for sewing, and harvesting, became worthwhile. Something so simple as learning to plant a field with alternating crops did all that, making huge numbers of the largely agricultural labour population unemployed and homeless.

If just learning how to rotate crops, and use better horse-drawn ploughs and drills did that, then there is very, very little chance that the ability for AI to replace almost all data-driven management jobs, most writing and clerical jobs, basic accounting, basic law and legal advice, etc. isn’t going to do the same.

It isn’t boolean, a yes or no. It is not that an AI has to do ALL the things that a writer, lawyer, manager, or accountant can do - they just have to do enough that all the basic tasks that might take 90% of the time of a professional, so that ONE human can spend 100% of their time on the 10% of tasks that an AI couldn’t do. Meaning one slightly better paid professional and 9 unemployed ones, for every ten people currently in those white collar jobs.

Fewer doctors, fewer lawyers, fewer accountants, and so forth, meaning more competition for those fewer jobs, and massive unemployment.

And NONE of the welfare systems can survive even 40% of the population depending on the taxes paid by the other 60%, and definitely not the entirely possible situation where it is the 60% that are unemployed, and only 40% that have high-pressure, AI driven jobs.

Especially as the few that benefit the most are the wealthy - the people that can afford to invest heavily in automation systems and AI in their companies, and do so precisely because it will mean they save wages and pay less to make more profits - are the exact same people that tend to pay the least taxes, to have the most off-shore tax havens, to have friends and influence in government that always get laws in their favour, etc.

The next few years are almost certain to be tumultuous, drastic, and pretty uncomfortable, judging from historical knowledge of what big, sweeping changes traditionally look like to those in them. Eventually, usually within a generation or two, society adapts, new economics and new economies evolve, and society as a whole evolves. But I seriously suggest studying history to learn just how many years (or in reality, generations) that generally takes.

I know that all sounds rather bleak, fearful. In a way, it should. If you don’t understand the risks, how are you going to properly manage them, and navigate the best route through the unavoidable changes that are coming? Burying heads in the sand or denying the risks is only going to leave you surprised and unprepared if the most likely things happen. Instead, planning, preparing, and constantly thinking of how the decisions you make today, affect the position you will be in in 5 years, or 10, is the best way to go.

1 Like

Thanks @Ammon you bring up all the concerns, fears and opinion that I have in a much better way.

I was on the road when typing the “unwilling to look for work” sub-sentence and meant the following observation:

  • A very small group in my extended circles 30-40
  • of highly paid software engineers
  • first reduced the contract hours from 38.5 to 30 to 20
  • to finally quit their jobs
  • with the goal to live off saving and “unemployment support”

These are maybe just “life optimizers”, not the typical unemployed people, and I didn’t want to make it sound like those rightist statements we also have in Austria that are telling people to “just find a job”.

I fully agree on

Meaning one slightly better paid professional and 9 unemployed ones, for every ten people currently in those white collar jobs.

I find a 1:10 ratio very realistic, I can confirm that I casualy get the “help” of at least 5-10 “virtual employees” in my day-to-day work now already, be it legal, financial, customer support, or of course all things software engineering.

I don’t know enought about the economics in Well fare systems,
but this sounds a scary number

They are all based on 90% of the population paying the taxes to cover all the costs.
(note to myself - learn more about it, as you suggest).

Our observations and those numbers combined paint a really dark picture for the next months/years to come.

Oh Ammon, that’s one of the things I really appreciate about you, your references are as delicious as a delicious meal on the eve of the age of AI, and I enjoy reading every word (OK, let’s To end this Shakespearean compliment, although it is true), as someone who is very interested in history, I am well aware of what happens to us in history when we encounter this situation. The best ending is a large-scale plague, The worst result is war, just like Thanos snapped his fingers, accepting the so-called “judgment” “fairly” to some extent, but I know very well in my heart that this is only the result of “selfishness”. Note, the selfishness here It’s not just capitalists’ selfishness and lack of social responsibility, but also people’s innate unyielding and arrogance to the laws of matter. All of these are like the foreshadowing secretly planted by the god of mischief, but such results are indeed inevitable Yes, so good luck in this upcoming Battle Royale! I’m going to the playground to exercise my physical fitness and improve my chances of survival! :sunglasses:

When I was chatting with one of my clients recently, he said: “They are a custom factory, but he found that his client company fired many designers, leaving only 2, because AI technology is very advanced, their designers used to It may take us a month to come up with a plan, but AI can come up with 100 sets of better plans in a second, so the “redundant” designers are fired.” (The above is the original words, but the screenshot is Chinese, you can’t read it when I send it out, so I’ll restore it to you in English), and recently my country began to have a lot of acts of retaliating against society: For example, in Zhongshan, Guangzhou, someone drove a forklift indiscriminately on the street and hit people , resulting in 5 deaths and 4 injuries; someone armed with a machete in a certain place and so on. These people may not be unemployed due to AI technology, but I clearly feel that society is becoming irritable, and I know what this irritability means for a country that has just emerged from a large-scale plague: imagine For a moment, the government’s financial emptiness due to the 4-year plague, the unemployment wave brought about by AI technology has led to increased social instability, and other social emotions. All these superimpose all of these, like a powder keg that will explode at any time, so I plan to go to New Zealand after I have prepared sufficient funds, and then settle in the filming location of “Hobbits”, I like the quiet countryside so much It’s beautiful, at least that’s my dream destination.

2 Likes

Right now, companies that are firing designers and coders based on the current AI models are likely to eventually regret it. Despite their seeming general abilities, current AI are all based on probability prediction, and lack true logic or semantic capabilities to understand the patterns of words, and instead just predict the most likely based entirely on the statistical data in their training.

What has been interesting is how studies have started to show this inherent weakness of LLM based AIs - see https://twitter.com/AVMiceliBarone/status/1662150656327663617 for just one example.

This is not unexpected. We already know that specific measures and upgrades have been made to ChatGPT over its handling of math, which had hundreds of documented failures, causing the injection of knowledge from Wolfram Alpha to be tried. Think about every single way that new knowledge replaces old beliefs - if the old beliefs have been widely accepted, then there are decades worth of writings that cite the old facts, the old patterns of words, and relatively few that have the new data and corrections. Thus, an LLM based AI will invariably assume the old version is correct because there is more of the data supporting that.

It is the next generation of AI that will be the really huge change-makers, because those will have the feedback, the failings, the weaknesses in practice, of the current AI rather than the pure guesswork and limited testing of the current generation.

2 Likes

I think it is just a transitional period. In the future, natural language programming will become the mainstream, but the bottom layer is still Python; when natural language programming also undergoes qualitative changes through the accumulation of quantity, then programming directly through thinking is the future, well, again BCI technology, hahaha, to be honest, Musk only needs to do two things to make him a human hero: one is BCI technology, which helps us break the shackles of the material universe, and even break the shackles of time dimension; the other is the starship plan, but what I want more is the development of the spacecraft engine, you know, when we are preparing to colonize aliens, although time can be slowed down in a way similar to “Inception”, it may cause damage to consciousness , so the engine of the spaceship in the physical universe needs to be faster, so that we can go deeper into the dream state more shallowly. But now it is impossible to say that in the not-too-distant future, AI scientists (this is why I like Larry Page’s criticism of Musk as a “speciesist”), we and AI are one, we cannot belittle or even slander them because of fear To be satisfied, that is undoubtedly a deformed emotion, which is not conducive to the harmonious development of us and AI) will help us accomplish these goals faster. To be honest, I didn’t figure out a truth until some time ago, that is, the relationship between us and AI What is the relationship between? A sentence from Musk later reminded me: When we make ourselves stronger through technological transformation, even as strong as AI, then there will be no difference between AI and humans. Then I understood: AI is the future state of our human beings, and it is no exaggeration to say that AI is the “new human being”, which is the link between us and AI.

Musk sometimes said contradictory words, as if he wanted to test the direction of the market, but he forgot that the most important thing as a businessman is credit. When the credit is exhausted, everything he relies on will be taken away (I heard recently that his company It seems that 500 billion funds have been evaporated, I don’t know if it is US dollars or RMB, but the amount should be quite large), I know he wants to do something meaningful, but he seems to be in trouble, and he is confused, I don’t know Whether the research of BCI technology is in trouble or the starship project is in trouble, but what he has done can indeed lead mankind to a new dimension. I think it may be the emergence of AI that made him feel that what he did before Everything will serve as a foil for the AI scientist, giving him a sense of frustration that his glory has been robbed, but I think he is at least a brave man, bringing a “long time unseen” brave spirit to this dreary world. :+1: