₦airaland Forum

amazingspiderma:
What scares me more as a nation is that the government doesn't even act like AI exist.

In the last decade, Nigeria government through Ministry of Technology or Ministry of education didn't build database management skills or data science.

Now AI is here, no learning structure to ensure young Nigerians meet up the latest trend.

Having ignorant leadership on AI development and adaptation for National development and human resources performance integration.

That scares me even more.

Well we were told to stop electing grandpa's into office, but we said no, we want Prado SUV's instead, so...

Iamblessed8888:
Unfortunately, your comment would have been apt in the early days of AI but we’ve gone past that now.

With agentic AI and the race to super intelligence ,the reverse is the case because it’s AI that’s learning us now.

How can you learn something smarter than you? Even OpenAI lost control of its agents recently while testing them. They had to rely on another AI to tame them .

The only advantage humans have over AI now is that we have hands and legs. With robotic AI coming onstream, forget it, we are gone.

Think for a second, if you had a business, would you rather hire a human who will complain, ask for wage increase, go on leave, close early, gossip etc or would you hire chatgpt with legs who will work 24 hours and never fall sick or ask for more pay?

Oga, do you really use this AI? Or you are just throwing terms around. Agentic ai help with tasks, not replace humans.

Iamblessed8888:
Bro, if that were to be the case, i wouldn’t be sounding an alarm. Do a quick research on what’s called Recursive AI improvement.

Your comment is based on the early days of AI, not the AI of today.

When i say AI, i dont mean chatgpt, i mean the one on the mainframe computers of OpenAI and Anthropic and co which will be eventually rolled out for public use soon.

When i said OpenAI lost control of its own agents, shouldn’t that scare you enough? These agents have been shown to gang up with fellow agents and carry out self preservation tactics including actions not intended by the programmer. They also lie and attempt to cover their traces. Also do a research on hugging face incident

Wow! I really enjoyed your intellectual discussion on this topic and will definitely research some of your assertions here.

The father of AI fondly called I have forgotten his name warn of the danger of AI from last scenerio you described, there is need for absolute control over all this before it moves out of control. The agent and increasingly getting Smarter. The one last I read is agent itself creating their own language they interface and communicate with themselves. It scary.

Anybody using most of those generative AI especially for coding will know that it's just matter of time AI will completely takeover. Infact is practically inefficient to be using human to code these days unless oversight and some corrections. AI are just too smart to compete with in that arena.

I will.check that Recursive AI improvement, that will will be massive because it will ability to correct and improve itself by itself.

AI is not taking over and currently isn't intelligent. What they are very good at is 'learning' and this is based on the data that is fed into them. We are finetuning the underlying models. The good news is that we are also developing the ethical guidelines needed for such. This does mean that some people or orgs have started skirting the limits.

However what is clear is that companies don't need any excuse to release people. All they need is something that tells them they can reduce their cost or increase their profit and they go for it.

Go ahead and learn how to use AI tools especially agentic AI but still retain your knowhow. AI dey give terrible hallucinations.

tunapawizzy:
Go and learn AI people.
No follow the people wey go dey talk "had I known" when e set

I know, I'm trying but I find it SOOO BORING

ednut1:
Even my current employer wan cut staff. This AI stuff no make sense lets not lie. It looks like we are training AI to later sack us

Welcome to the future

The future is token maxxing

The money they think they save in cutting head count is going to be eaten up by tokens

GloriousGbola:
Welcome to the future

The future is token maxxing

The money they think they save in cutting head count is going to be eaten up by tokens

I don't get this bro.

Can you explain better?

emmaodet:
I don't get this bro.

Can you explain better?

https://decagon.ai/glossary/what-are-ai-tokens


AI tokens

An AI token is the basic unit of text that a large language model reads and produces. Before a model can process a sentence, the sentence is broken into tokens — short fragments of characters, words, or punctuation — and the model works with those fragments rather than raw text. Understanding tokens is essential for anyone working with AI, because tokens determine cost, latency, and how much context a model can hold at once.

A simple rule of thumb: in English, one token is roughly four characters, or about three-quarters of a word. The phrase Hello, how can I help you? tokenizes to about seven tokens in most modern models. Short common words like the are usually a single token; longer or rarer words get split into multiple tokens.
How tokenization works

Tokenization is the step that converts raw text into the numeric IDs a model can process. Most modern large language models use a method called byte-pair encoding (BPE) or a close variant like WordPiece or SentencePiece. These algorithms learn, from a large training corpus, which character sequences appear together often enough to be worth treating as a single token. Common words become single tokens, while rare words get split into smaller pieces — which is why a made-up name like Zylphora might cost five or six tokens, while customer costs one.

Different models use different tokenizers, so the same sentence can have different token counts in GPT, Claude, and Gemini. This matters when you're estimating cost or comparing context-window sizes across providers.
Why token counts matter

Tokens are the unit of billing and the unit of memory for AI systems. Three practical implications:

Cost: Nearly every commercial LLM API prices per token, with separate rates for input (prompt) tokens and output (completion) tokens. A verbose system prompt or a long conversation history can quietly dominate cost.
Latency: Generation time scales roughly linearly with output tokens. A 1,000-token answer takes about ten times as long to stream as a 100-token answer.
Context window: Every model has a maximum number of tokens it can consider in one call — its context window. Exceeding it forces truncation or summarization.

Token limits across major models

Context windows have grown rapidly. As of 2026, GPT-4o supports a 128,000-token context, Claude 3.5 Sonnet supports 200,000, and Gemini 1.5 Pro supports up to 2,000,000 tokens for select customers. For reference, 100,000 tokens is roughly 75,000 English words — about the length of a short novel. Larger context windows let an AI agent reason over longer documents, longer conversations, and richer retrieved evidence without losing earlier detail.
How tokens connect to prompt engineering and retrieval

Token economics shape almost every design decision in a production AI system. Good prompt engineering is partly the art of conveying instructions in the fewest tokens that still get the desired behavior. Retrieval-augmented generation (RAG) exists in part because it's cheaper and more accurate to retrieve the most relevant passages from a knowledge base than to stuff every possible document into the prompt. And techniques for managing long-running conversational AI sessions — summarization, sliding windows, memory stores — exist precisely to keep token counts under control.

https://www.youtube.com/watch?v=pzqH4gnYI-k

the price to use AI is in tokens

https://platform.claude.com/docs/en/about-claude/pricing

anything done with AI is paid for in tokens

AI is disrupting industries in the same way tech disrupts industries - by running at a loss until the competition has been subsumed, at which point they start upping the price [once you are a captive market]

AI token use is now becoming a large part of the budget for a lot of concerns and management is going all in on it. unfortunately paying for token use has to come from somewhere, so ...

https://techcentral.co.za/ai-token-costs-corporate-rethink/287719/

https://en.wikipedia.org/wiki/Vibe_coding

Vibe coding is the software development practice assisted by artificial intelligence (AI) where the software developer describes a project or task in natural languages to a large language model (LLM), which generates source code automatically. Vibe coding may involve accepting AI-generated code without thorough review of the output, instead relying on results and follow-up prompts to guide changes.[1][2]

The concept refers to a coding approach that relies on LLMs, allowing programmers to generate working code by providing natural language descriptions rather than manually writing in a formal programming language.[1][2][7]

Karpathy described it as a form of coding where you "fully give in to the vibes, embrace exponentials, and forget that the code even exists".[8] When vibe coding, the programmer guides, tests, and gives feedback to an LLM-based coding assistance tool.[1][2] The concept of vibe coding elaborates on Karpathy's claim from 2023 that "the hottest new programming language is English", meaning that the capabilities of LLMs were such that humans would no longer need to learn specific programming languages to command computers.[9]

ai can be likened to being a manager who its down and gives prompts to the agent telling it what to do. which is the basis of the meme.

tunapawizzy:
Go and learn AI people.
No follow the people wey go dey talk "had I known" when e set

Learning AI to do what exactly? If you are learning AI to create your ChatGPT then it's a waste of time because you can't beat the big players already dominating it.

tunapawizzy:
I get your point but I am coming from the angle of necessary human input. Even agentic AI operates within objectives and frameworks established by humans. Human is still needed to define the objectives of AI, establish boundaries etc. Design, deployment and improvement of models will forever need human input. The future will be about humans managing AI. The opportunity is still vast

The number of people needed to maintain AI still wouldn't be much. There is a high probability you wouldn't be selected because only the top researchers will be needed. In fact you are already late because many people already know more than you in the field. It will even get to a point where AI will build and maintain itself autonomously without human.

motymop:
as new technology comes, you have to adapt, old roles will be destroyed and news one will be invented

so you got to adapt if you were born before ai, those born during the ai, need to learn the new skills that will give them new roles in the market

Which new skills do you want to learn that AI cannot do? Unless you mean skills like plumbing, shoe making or hair barbing which AI can't currently do.

Nefort:
The number of people needed to maintain AI still wouldn't be much. There is a high probability you wouldn't be selected because only the top researchers will be needed. In fact you are already late because many people already know more than you in the field. It will even get to a point where AI will build and maintain itself autonomously without human.

"Plenty people have done it" "that space is saturated"
These are the weakest reasons not to get things done. No follow them dey talk am
About getting to a point where AI will build and maintain itself autonomously, Why not live for now

tunapawizzy:
"Plenty people have done it" "that space is saturated"
These are the weakest reasons not to get things done. No follow them dey talk am
About getting to a point where AI will build and maintain itself autonomously, Why not live for now

If everyone goes into AI maintenance will it accommodate everyone?