The State of AI - September 2025

"AI could wipe out half of all entry-level white-collar jobs — and spike unemployment to 10-20% in the next one to five years." -- Dario Amodei, Anthropic CEO

The State of AI is changing rapidly. What does the future hold?

The State of AI Today—and What It Means for Our World

By Phil

Artificial intelligence has moved beyond the realm of science fiction and into everyday reality. It now recommends the music we hear, helps farmers predict crop yields, assists doctors with early diagnoses, and writes code in the middle of the night.

The question before us is no longer whether AI will shape our world. The question is how will it do so?

A Force That’s Reshaping the Economy

AI adoption has exploded. More than three-quarters of organizations now use it, and global investment in generative AI alone climbed past $30 billion last year. Some economists see echoes of the Industrial Revolution, predicting annual GDP growth in the double digits.

But it’s not a straight line to prosperity. Recent studies show developers sometimes code slower when using AI, and workers around the world feel the pressure of jobs that may be automated away. By 2025, up to 16% of roles could be displaced, with only partial replacement by new ones.

Behind these numbers are people wondering if their careers—and communities—will still thrive in an AI-driven economy.

According to Dario Amodei, CEO of Anthropic, he expects many frontline and entry level jobs to be taken by AI within 5 years. The truth is, it is happening already. Many CEOs have performed large scale layoffs with the often misguided belief that AI can replace humans.

AI can replace humans, but it still has some ways to go before it can really do it en masse. There will be jobs that will be harder for AI to replace, but the ultimate aim of corporations and CEOs is to be able to replace humans as much as possible, as soon as possible, to bolster their capitalistic desires. They don't care about the impact it has on the worker force.

I work in AI for my day job, and I've seen how AI is already being used to automate tasks from reception duties, to parsing large amounts of documentation and collating it. It is already happening. It remains to be seen the best way to remain relevant in the coming years, but as I've covered in an earlier article on SidLinx, there are things we can do. Look out for an updated version of that article to provide you with some useful information to keeping yourself relevant in the coming AI powered economy.

And if it makes you feel any better, the CEOs doing all the firing might soon find themselves on the unemployed list. In the US, CEOs are paid stupid amounts of money for work that can often be replaced by AI. As AI gets smarter, the CEOs will be next to go.

A Double-Edged Tool

AI is a paradox.

• It can spot early signs of a hurricane, saving lives.

• It can also generate deepfakes so convincing that public trust itself begins to crumble.

This “liar’s dividend”—where even real events are dismissed as “probably AI-generated”—threatens the fabric of democratic conversation.

And then there’s the environment. Training a large model can emit hundreds of tons of CO₂. Experts warn AI’s energy demand could soon rival that of a small nation. Yet at the same time, AI is being used to design greener energy grids and predict climate risks.

The trouble with this is it depends on the human controlling the AI to know whether the outcome is going to be ultimately helpful to society as a whole, or the individual/corporation wielding it. AI has the power to do great things, but if you look at the US AI industry, they're not doing it to better humanity, they're doing it to become the new technocracy. I have little desire to be a serf of the tech bros. Most of them I have little respect for.

You have open source AI that is starting to really start challenging frontier models, and that is where you have the great equalizers. Democratizing AI by putting it in the hands of all, instead of the 1% is how AI can become a real tool for good. Think of AI as an amplifier of the party wielding it.

A Crossroads of Power

Governments are waking up.

• The European Union’s AI Act sets one of the world’s first comprehensive frameworks, banning the riskiest uses.

• France has pledged €200 billion in AI investment as a matter of national strategy.

• In the U.S. and China, AI has become a geopolitical race—about more than software, but sovereignty itself. The US is chasing capitalism and China is chasing open source.

As a product of parents who stupidly fell for the neoliberalism con, I definitely am opposed to frontier AI being only in the hands of the powerful. Those trickle down economics have been 'any day now' for about 40 odd years. Sid may know more about that.

Tech leaders are equally divided. Some predict artificial general intelligence within the decade, freeing humanity from work. Others warn superintelligence could be our last invention if it slips beyond control. Both outcomes pose great risks. 'Freeing' humanity from work raises so many questions. Whether people like it or not, the AI genie is out of the bag, and it is up to responsible humans to ensure the better outcome of AI.

Choosing Our Direction

AI will not decide for us. It reflects the intentions we build into it. If we pursue only profit, the outcome may be inequality and disinformation. If we treat it as a shared project, it could enrich education, health, sustainability, and culture.

The heart of the matter isn’t technical—it’s human. Will we teach students to think with AI rather than let it think for them? Will we design systems that preserve dignity instead of reducing people to data? Will efficiency gains be shared widely, or hoarded narrowly?

Moving Forward

The story of AI is still unfolding. Every policy passed, every classroom conversation, every workplace experiment adds to the shape of what’s coming.

“Technologies this powerful don’t bend toward justice on their own. They bend according to the values we embed in them, and the courage we summon to govern them.”

The future of AI is not fully known. Experts like Dario Amodei can speculate. Joe Average can fear doom. The main thing is to be sensible and level headed. I always say there's one ace we have up our sleeve if we're smart enough. Have someone always handy to pull the power plug if SkyNet rears their ugly head. However, the opportunity for leaps humanity cannot comprehend, such as cracking the code for curing cancer and all ailments within 15 years is not to be sneezed at. The opportunity—and responsibility—is ours: to make sure the tools we build remain firmly in the service of humanity and not to those few at the top.