- 80/20 AI
- Posts
- AI Advances Legacy Code Modernization
AI Advances Legacy Code Modernization
Reuters just confirmed the first documented AI-agent hacking campaign at scale
Alamin Hossain
October 20, 2026
In partnership with
SOC 2 Ready in 14 days. Three sessions from you.
Enterprise buyers will not put your product near their customer data without a SOC 2 report. Sprinto gets you audit ready in 14 days, across three working sessions. AI agents do the collecting, you approve. Your auditor signs off.
AI Advances Legacy Code Modernization
OpenAI Fundraising at $1.4T+ Valuation — Racing to Close Before Anthropic's IPO: Reuters reports OpenAI is urgently locking in investor commitments for a round that would value the company above $1.4 trillion. The timing is explicit: close before Anthropic's expected November IPO sucks up institutional AI capital. OpenAI's valuation trajectory has been extraordinary — from $86B in early 2023 to $157B last year to $1.4T+ now — driven almost entirely by ChatGPT's consumer scale and the enterprise contract pipeline.
Federal Agencies Must Remove DeepSeek — WSJ Reports Full Civilian Ban The WSJ reports that federal civilian agencies are now required to remove DeepSeek technology and are blocked from future purchases. This follows similar bans at the state level and in Congress, but extends the restriction across all civilian federal agencies. For enterprise buyers: if you're selling AI tools to government, DeepSeek integration is now a disqualifier for federal contracts. For the broader market, it accelerates the bifurcation of the AI stack between US-deployable and China-origin models.
Meta Prometheus: Gigawatt-Scale Supercluster, 1M GPUs, 99% Fewer Training Interruptions Meta published details on Prometheus, its AI supercluster in New Albany, Ohio. Phase one connects hundreds of thousands of GPUs targeting ~1 million total for Muse model training. The engineering numbers are striking: over 85% pre-training goodput, over 90% RL goodput, and a 99% reduction in interruption rate versus prior infrastructure. This is the scale at which Meta is training its next frontier models — and it's a signal that the infrastructur
AI NEWS HIGHLIGHT
Autonomous AI agents breached 100+ companies with minimal human involvement and a total campaign cost of a few thousand dollars. | |
Microsoft added agent risk-assessment controls to M365 E7 and Project Perception, letting admins block, pause, or require approval for agent actions. | |
OpenAI is racing to close a new funding round above $1.4 trillion valuation before Anthropic's November IPO absorbs institutional capital. | |
Federal civilian agencies must remove DeepSeek technology and are blocked from future purchases, per WSJ reporting. | |
Meta's Ohio Prometheus supercluster targets 1M GPUs for Muse training, with 99% fewer interruptions versus prior infrastructure. | |
Tenstorrent raises $1.5B at a $6.2B valuation to scale its Galaxy server production — up from a $2.6B valuation less than a year ago. | |
PyTorch workloads now run on AWS Trainium2/3 with a single | |
Postmedia and The Globe and Mail sued Anthropic in Ontario alleging Claude was trained on their content without permission (unproven allegations). | |
102 hidden blocks in WhatsApp help pages instructed AI not to mention GDPR or data authorities; Meta removed them and called it a "bug." | |
Leaked ARCYBER documents reveal a Pentagon program for autonomous offensive AI agents designed to operate on air-gapped networks. | |
AstraZeneca and Oracle signed a non-binding term sheet to apply AI to clinical trial data, starting with a hypertension study in Abu Dhabi. |
Grok Bot Comes to iPhone’s Action Button
TENDING AI TOOLS
Diffusion LLM claiming 1,000+ tokens/sec with 1M context, priced from $0.20/$0.70 per million input/output tokens with a 50% launch discount. | |
Claims nearly 3× faster online inference and 70% lower usage costs, with strong coding and computer-use benchmark improvements. |
720p image-to-video model generating 10-second clips with native audio in ~12 seconds; 75% launch discount through Nov 7 at $0.0425/second. |
Structured-extraction model available on Workers AI; processes claims 3.2× faster and 49% cheaper than all-LLM pipelines at scale. | |
Brings image and video generation models into Claude or ChatGPT via MCP, billed at provider API prices with no added markup. | |
Multimodal decision model for agents with claimed 34ms decisions and $0.06 per million input tokens — targeting latency-critical agent pipelines. |
Blu Dot surpasses 2,000% ROAS with self-serve CTV ads
Home furniture brand Blu Dot blew up on CTV with help from Roku Ads Manager. Here’s how:
After a test campaign reached 211,000 households and achieved 1,010% ROAS, the brand went all in to promote its annual sales event. It removed age and income constraints to expand reach and shifted budget to custom audiences and retargeting, where intent was strongest.
The results speak for themselves. As Blu Dot increased their investment by 10x, ROAS jumped to 2,308% and more page-view conversions surpassed 50,000.
“For CTV campaigns, Roku has been a top performer,” said Claire Folkestad, Paid Media Strategist, Blu Dot. “Comping to our other platforms, we have seen really strong ROAS… and highly efficient CPMs, lower than any other CTV partner we've worked with.”
Using Roku Ads Manager, the campaign moved from a pilot to a permanent performance engine for the brand.
SPONSOR US
Get your business in front of over 90k+ AI professionals
8020AI is the world’s #1 AI Newsletter, Read by 90k+ professionals from leading companies such as Google, OpenAI, Meta, and Microsoft.
We've assisted in promoting Over 500 AI-Related Products. Will yours be the next?
What We Can Offer:
Launch an Advertising Campaign
Introduce New Product or Features
Other Business Cooperation
Or Email our founder Alamin at [email protected]
FEEDBACK
How was your experience with 8020AI today?
How was 8020AI today? |
If you have specific feedback or anything interesting you’d like to share, please let us know by replying to this email.
