AImpact

Sharing the pulse of AI and LLMs. Breaking down news, highlighting innovation, and championing a future where tech mirrors the best of Star Trek—open, ethical,

[Wes Roth @WesRoth]

I can't believe how good AI video is getting...

we're so cooked, I could not stop watching this

[AImpact @JLockesGhost]

@WesRoth That shouldn't be that good! Thats Amazing, cant wait to get my hands on tools like this!

[AImpact @JLockesGhost]

@grok @Jimmy_Rockford @TansuYegen @Grok Who is she

[AImpact @JLockesGhost]

@tonbendery @MarioNawfal Comparatively, almost 0. People kill thousands a year.

[Robert Youssef @rryssf]

Holy shit... this might be the next big paradigm shift in AI. 🤯

Tencent + Tsinghua just dropped a paper called Continuous Autoregressive Language Models (CALM) and it basically kills the “next-token” paradigm every LLM is built on.

Instead of predicting one token at a time, CALM predicts continuous vectors that represent multiple tokens at once.

Meaning: the model doesn’t think “word by word”… it thinks in ideas per step.

Here’s why that’s insane 👇

→ 4× fewer prediction steps (each vector = ~4 tokens)
→ 44% less training compute
→ No discrete vocabulary pure continuous reasoning
→ New metric (BrierLM) replaces perplexity entirely

They even built a new energy-based transformer that learns without softmax no token sampling, no vocab ceiling.

It’s like going from speaking Morse code… to streaming full thoughts.

If this scales, every LLM today is obsolete.

[AImpact @JLockesGhost]

@NVIDIAAIDev When will it be available and Price?

[AImpact @JLockesGhost]

@RadioGenoa Keep Fighting!

https://t.co/mroa6Lxvl2

[AImpact @JLockesGhost]

Fight the good Fight...

[AImpact @JLockesGhost]

Fight the good Fight...

[AImpact @JLockesGhost]

@elonmusk I don't like the extra pay tier for grok, but other than that, I'm highly impressed with its speed of development.

[AImpact @JLockesGhost]

I think Grok is the faster moving LLM/AI. What do you think?

[Elon Musk @elonmusk]

The Grok Singularity

[AImpact @JLockesGhost]

🚨 News / Alert

🚨 OpenAI just launched ChatGPT Pulse — a proactive AI that doesn’t wait for prompts, but sends you daily updates.
The Verge: https://t.co/QHu1QR7iuy

For now: Pro users only. But this feels like a turning point.

[AImpact @JLockesGhost]

@MurdaMooky @libsoftiktok You mean the stores off the interstate with the giant fucking billboards for miles and miles, those stores....

[AImpact @JLockesGhost]

Healthcare’s AI Gold Rush
AI’s set to pump $187B into healthcare by 2030—diagnosing faster than your doc! 🩺 Imagine AI catching your cold before you sneeze. Miracle or meddling? Health nuts, drop your take! 💉
#AI #HealthTech #DocBot
🩹 Like for smarter checkups, repost to cure the feed, follow for wellness wins!

[AImpact @JLockesGhost]

#BREAKING 🔥
Tylenol + Pregnancy — Evidence Emerging, Risks Still Uncertain

🚨 FDA has begun updating acetaminophen (Tylenol) labeling, citing studies that suggest possible associations with autism & ADHD when used during pregnancy. Still, scientific consensus remains: no proven cause-and-effect.

🤖 Where AI comes in:
• It can amplify misunderstandings if media/algorithms present early findings as fact.
• It also holds the power to analyze large datasets, weigh conflicting studies, and help public & practitioners make informed choices.
• Guardrails needed: accuracy, transparency, distinguishing “possible link” vs “proven outcome.”
📝 Bottom line: Pregnant folks should talk to their doctors. Policymakers & tech platforms should help ensure messages are responsible & evidence-based.

🔗https://t.co/ybHvJepexx

#Tylenol #Health #AI #DataScience #ScienceCommunication

[AImpact @JLockesGhost]

#BREAKING 🔥

U.S. Secret Service just dismantled a massive telecom threat around UNGA: over 300 SIM servers & 100,000 SIM cards seized within 35 miles of NYC.

Why this matters for AI / LLM / tech watchers:
Such networks could be leveraged by malicious actors to feed AI-driven disinformation or spam systems, or to evade automated content/filtering moderation.

Encrypted + anonymous communications powered by this infra could hide coordinated manipulation—botnets + AI combined become much harder to track.

Defense systems (AI/ML) must step up: anomaly detection, synthetic-traffic detection, identity verification of message senders.

More here: https://t.co/DzlFyA6CWO

[AImpact @JLockesGhost]

TRENDING: Digital ID + AI — What’s at stake

Digital identity systems powered by AI are rapidly becoming the backbone of how governments and enterprises verify who we are. But there’s tension:

✔️ Upsides: faster onboarding, reduced fraud, better access to services.
✔️ Risks: deepfakes & synthetic identities, bias in biometrics, misuse of personal data, and gaps in regulation.
To build systems people can trust, we need:
• Strong legal protections for identity data and consent.
• Transparency: what data is collected, how it's used, who sees it.
• Inclusive design so no one is left behind (people without tech access, etc.)
• Technical safeguards: anti-spoofing, verification & detection of synthetic content.

We’re at a crossroad — either AI makes identity verification more just & secure, or opens doors to abuse.
#DigitalID #AI #Privacy #TrustTech