"Today's AI is tomorrow’s dial-up!" - Futurist Jim Carroll

There’s a phrase I like to use: “Today will involve the slowest rate of change for the rest of your life!”

That concept seems quaint in the era of fast-moving AI. In fact, we should rephrase it: “By the time you finish reading this sentence, the state-of-the-art has already moved on.”

Think back to 2024. We were impressed by a chatbot that could summarize a PDF. For those of us of a technical bent, we thought a 128k ‘context window’ was a “breakthrough.” In retrospect, that was the screeching, static-filled noise of the 56k modem era. We were in the AOL-era of AI!

Today? Welcome to the broadband era of intelligence.

In just the last 24 months, the landscape hasn’t just shifted. It’s undergone a seismic shift. The ground we knew below our feet has been pulverized and rebuilt. Think about just a few of the AI trends as we go into this new era:

The Memory Explosion: We’ve graduated from "summarizing a document" to "uploading the entire library." Leading models like Gemini now process millions of tokens at once, compared to the 128k context windows we were celebrating as breakthroughs just 18 months ago. We aren't just prompting; we’re giving AI an entire War & Peace document set to churn through.

The Price Collapse: Intelligence is now a commodity utility. The cost to run GPT-4-class reasoning has plummeted by 98%. What cost $60 is now essentially "too cheap to meter" at less than $0.75.

From "Chatting" to "Doing": 2024 was about talking to a screen. 2026 is about Agentic AI. We're moving past chatbots to autonomous agents that negotiate, navigate CRMs, and execute project workflows without being babysat.

The last few weeks have had me working with Claude Cowork and Claude Copilot, and my entire concept of AI has changed. I don’t just ask it questions — I now instruct it to go off and do the work on my behalf. That’s a subtle but important change, and once you wrap your head around what that change brings, your head explodes.

And I haven't even yet jumped into the world of Claw! It's on the list!

It is evident that not only is the power and capability of AI accelerating, but we are rapidly building a new operating system around it. If you are still "playing around" with prompts, you’re still waiting for the handshake signal on a dial-up connection. You're stuck with a modem connection when the rest of the world has installed fiber.

The bandwidth of possibility has expanded. The question isn't whether the tech is ready. The question is: Are you ready for the speed of the "Broadband" era, or are you still waiting for the page to load?

----

Futurist Jim Carroll started his online experience with a 300-baud modem in 1982. Think about that.

**#AI** **#DialUp** **#Broadband** **#Acceleration** **#Technology** **#Future** **#Change** **#Innovation** **#Agents** **#Claude** **#Speed**

Original post: https://jimcarroll.com/2026/03/decoding-tomorrow-daily-inspiration-the-acceleration-of-ai-todays-ai-is-tomorrows-dial-up/

Im Sorry, I wasn’t aware it was that easy.

I’m sure there is more in the article. But seeing that picture had me #ROFL (metaphorically).

Timeline: Tech is About to Accelerate Because AI Scales Itself https://web-strategist.com/blog/2026/03/15/the-next-tech-waves-are-accelerating/

#AI #acceleration #exponential #genAI

No Way Out Within Capitalism -->

"One of the great hopes for artificial intelligence—at least, among workers—is that it will ease workloads, freeing people up for more high-level, creative pursuits. So far, the opposite is happening, new data show.

In fact, AI is increasing the speed, density and complexity of work rather than reducing it, according to an analysis of 164,000 workers’ digital work activity. The data, from workforce analytics and productivity-tracking software company ActivTrak, covers more than 443 million hours of work across 1,111 employers, making it one of the biggest studies of AI’s effects on work habits to date.

Examining AI users’ digital activity 180 days before and after they began using such tools on the job, ActivTrak found AI intensified activity across nearly every category: The time they spent on email, messaging and chat apps more than doubled, while their use of business-management tools, such as human-resources or accounting software, rose 94%.

Meanwhile, the amount of time AI users devoted to focused, uninterrupted work—the kind of concentration often required for figuring out complex problems, writing formulas, creating and strategizing—fell 9%, compared with nearly no change for nonusers.

“It’s not that AI doesn’t create efficiency,” said Gabriela Mauch, ActivTrak’s chief customer officer and head of its productivity lab. “It’s that the capacity it frees up immediately gets repurposed into doing other work, and that’s where the creep is likely to happen.”

Such habits aren’t exactly what AI evangelists have predicted. A number of tech and business leaders, from Bill Gates to JPMorgan Chase’s Jamie Dimon have suggested that AI could ultimately lead people to work less, not more, and result in a shorter workweek. Elon Musk has said that, within 20 years, advancements in AI and robots could even make work optional."

https://www.wsj.com/tech/ai/ai-isnt-lightening-workloads-its-making-them-more-intense-e417dd2c

#AI #GenerativeAI #Automation #Productivity #Acceleration

Mikhaíl Ilín (@ilin_pt)

LLM 가중치를 실리콘에 직접 넣어 기존 범용 GPU 기반 추론보다 1000배 빠르다는 주장과 매우 빠른 데모를 소개하며, 'weights in silicon' 방식의 하드웨어 가속이 큰 성능 혁신을 가져올 수 있음을 강조합니다.

https://x.com/ilin_pt/status/2029931657369674101

#llm #inference #hardware #silicon #acceleration

Bergson, la durée et l’accélération technique : quand le temps vécu s’efface

Pour Bergson, la durée – le temps vécu, subjectif – s’oppose au temps mesuré des horloges. L’ accélération technique moderne écrase cette durée, sacrifiant la maturation et la mémoire profonde au profit de l’instantanéité. #Bergson #temps #accélération #mémoire #philosophie #temporalité

https://homohortus31.wordpress.com/2025/07/13/bergson-la-duree-et-lacceleration-technique-quand-le-temps-vecu-sefface/

Parakeet.cpp – Parakeet ASR inference in pure C++ with Metal GPU acceleration

https://github.com/Frikallo/parakeet.cpp

#HackerNews #ParakeetCpp #ASR #CPlusPlus #MetalGPU #Acceleration #HackerNews