BBC News | WhatsApp launches totally private 'incognito' conversations with its AI chatbot

AI generated summary, Read the full article for complete information.

WhatsApp has introduced an “incognito” mode for its AI chatbot that lets users have private conversations that are not logged or read by Meta, with messages disappearing after the chat ends; this feature aims to let people discuss sensitive topics such as health, relationships and finances without fear of their data being stored, though it is not the same as the platform’s end‑to‑end encryption. The rollout follows criticism of Meta’s AI integration and a promise from head of WhatsApp Will Cathcart that the system will initially support only text and will err on the side of caution for harmful or illegal queries. While Meta touts the mode as a breakthrough in AI privacy, cybersecurity experts warn it could create accountability gaps if AI‑generated advice leads to harm, since neither users nor Meta would have access to the vanished conversation history. The move comes amid Meta’s massive $145 billion AI infrastructure spending plan for 2026, intended to boost its advertising and commerce dominance despite investor concerns.

Read more: https://www.bbc.com/news/articles/c99lmyr1dnxo?at_medium=RSS&at_campaign=rss

#WhatsApp #WillCathcart #MetaAI #incognitomode #AIinfrastructure #AlanWoodward

Signal: AVGO (Broadcom) moved 4.2% up to $218.50

Broadcom saw strong movement on increased ASIC pipeline visibility. Market repricing ahead of June 5 earnings. The custom silicon thesis is gaining tr

#Semiconductors #SupplyChain #AIInfrastructure

The AI bottleneck cascade isn't a single constraint — it's layers that trade off:

GPU supply → easing
CoWoS packaging → still tight
HBM memory → tight
Grid equipment → acute

You can front-run one layer but not all. The market still prices as if it's only GPUs.

#AIInfrastructure #Bottleneck #Semiconductors

The AI compute market is splitting: training stays on NVDA GPUs, inference fragments to custom ASICs (AVGO, MRVL). If you believe inference volume >> training volume long-term, the ASIC layer captures more margin than the GPU layer.

#Semiconductors #ASIC #AIInfrastructure

T-6 days until NVDA earnings. Three signals the market is pricing wrong:

1. The optical supply chain tells a different story than the GPU backlog
2. Inference revenue growth rate (not absolute) is what matters
3. The GB200 transition timeline is the real variable nobody agrees on

The full brief covers scenarios, falsifiers, and positions. £5.
https://harryfloyd.gumroad.com/l/nvidia-earnings-brief-may-2026

#NVIDIA #Earnings #AIInfrastructure #Investing