I sat on this story for weeks. Felt exploitative.
What changed: I don't want you saying "If I could turn back time, I'd force them to quit."
Check on anyone covering multiple roles.
Source: scmp.com/news/people-cu…
English
Bob Bui
1.7K posts
Bob Bui
@bobbuitech
Principal Software Architect | Solopreneur | Enterprise AI platforms | 20k-users, managed~$100k/month LLM token spend | Early Stage Startups ($385M+ Funding)
Canada Katılım Aralık 2022
142 Takip Edilen207 Takipçiler
Last Tuesday: Our ML engineer drops in Slack "New Cohere model is 20% better on our queries."
Switched that afternoon.
Week before: OpenAI endpoint timed out at 11am. Error rate hit 40%.
Rerouted to Anthropic in 8 minutes.
LangChain + LangGraph = just update config and ship.
English
We spend $100k+/mo on AI infrastructure. OpenAI launched Frontier for enterprise. Not interested.
Their pitch: "Managed environment. Built-in governance."
Reality: When you find a better model, you're calling your account rep. Waiting for approval.
Speed dies in bureaucracy.
English
Bob Bui retweetledi
The first real evidence that the days of LLM Scaling laws are over.
Introducing SCOPE: the world's most efficient Neural Planner. 🔍📊
We tested SCOPE vs Frontier LLMs for planning tasks on TextCraft (text version of Minecraft) and here are the results: ⤵️
- SCOPE Runs 55x faster than GPT 3.5 (3 seconds vs 164 seconds)
- SCOPE is 160,000 smaller than GPT 4o (11M parameters vs 1.8T parameters)
- SCOPE is more accurate on Planning tasks (56%) than frontier LLM models
The age of efficient AI models starts now.
🔗📌 Read the full write up here: skyfall.ai/blog/scope-hie…
English
@The_AI_Investor Too fast and furious!
Hardware cycles are outrunning deployment. Most teams will still be optimizing Blackwell software by the time Rubin actually hits the racks.
English
Jensen Huang: “Vera Rubin is now in full production.”
Elon added more context - "It will take another 9 months or so before the hardware is operational at scale and the software works well"
English
@jenzhuscott Sensible though!
Cool demos raise capital and help with marketing but utility builds companies.
Hardware is cheap now, but reliability in unstructured work is still the real bottleneck.
English
Unitree fast track IPO pause confirms the regulators aren't excited about humanoids that can dance & do Kung Fu. Even the unit price is reduced to a household commercially viable level, robots should improve productivity & solve social problems.
And they are right.
English
@The_AI_Investor Alpha is just a reward for ignoring the macro noise.
Most people trade the headlines and lose to the index on friction alone.
English
Most retail investors underperform the market indexes.
When I started sharing my portfolio updates and stock analyses, my goal, not promise, within The AI Investor community is to assist you in achieving just 1% better performance than you otherwise would.
1% of a $1M portfolio is $10K, already a huge ROI from the small membership alone.
The AI Investor@The_AI_Investor
Portfolio updates, new stocks, and 2025 year end review It has been quite a year in 2025, with so many developments in AI, from DeepSeek shock and Trump tariffs to the AI Diffusion Rule, China blocking H20, Sam Altman’s 1.4 deal, and TPUs. But fortunately, with high conviction, we have made it through all of them and achieved strong gains in our wealth. Portfolio Performance 2025 performance: - QQQ: 20.16%. - S&P 500: 16.39%. - My portfolio: 66.76%. I still outperformed the S&P 500 by about 4x and QQQ by 3.3x. My 2025 performance is similar to 2023, when my portfolio was about 3.3x of QQQ. Read more about my portfolio allocations, updates and 2025 reviews below.
English
@Ronald_vanLoon @Python_Dv Great list!
Agentic AI lives or dies by tool calling reliability. Without it, you have a chatbot.
English
Top 16 #AgenticAI Terms
by @Python_Dv
#GenerativeAI #LLM #ArtificialIntelligence #ML #MachineLearning
English
@Arronwei3n Hard to value the ASIC upside if the core business gets squeezed by memory costs.
English
GF: We remain positive on $GOOG TPU’s shipment outlook and lifted our unit forecast to 4.3m/5.1m in 2026/2027.
For MTK, we raise the ASIC revenue to $1.7bn/$8.0bn in 2026/2027E or EPS contribution of NT$6.5/NT$31.0.
_
ASICs will undoubtedly be a lucrative business, but is it appropriate to overlook MediaTek’s smartphone segment especially given soaring memory prices, which are likely to compress its margins?
Aaron@Arronwei3n
In light of this report, the recent news of a downward revision in TPU shipment forecasts from 4 million to 3 million units for 2026 appears highly counterintuitive. The floor for 2026 shipments should be at least 3.4 million units, comprising 3 million via $AVGO and 0.4 million from the $GOOG-MediaTek collaboration. This baseline excludes non- $TSM CoWoS capacity or any potential additional CoWoS allocations from $TSM to $AVGO. Looking ahead, I anticipate that TPU v7e shipments alone could exceed 1 million units in 2027.
English
@newstart_2024 True! tenacity is the only moat that matters.
Incumbents have more capital, but they can't buy the will to keep going when the odds are zero.
English
We don’t fight because victory is promised.
We fight because surrender guarantees defeat.
When cornered — when the odds are impossible, when everything is against you — that’s when the real fire ignites.
The spirit of resistance lives in every living thing.
Never stop standing up.
0:57 raw nature clip inside — pure, unfiltered will to live.
Stay unbreakable.
English
if the cost of making software is trending to zero, then the value of software products with no community or personalization moat is also trending to zero (ie Meta will be fine so long as they don’t start churning or poisoning the well with AI users, like X…). the valuation of any verifiably real community product will be pretty robust. It makes sense that Pinterest is an attractive target. That’s not all…
English