Suman Deb Roy

I design AI systems. Previously with @buzzfeed, @betaworks, @columbiajourn, @MSFTResearch

[Suman Deb Roy @_RoySD]

Decoupling research from institutional overhead has such far reaching effects

[MTS @MTSlive]

SITUATION DETECTED: The White House Office of Science and Technology Policy has released a report directing federal research funding toward individual scientists and AI use rather than universities, per WSJ.

The directives will shape about $200B in annual federal R&D spending.

[Sam Dumitriu @Sam_Dumitriu]

It was smart of Tuchel to rest Saka against Argentina so he could be fully fit for the third place playoff.

[Suman Deb Roy @_RoySD]

Tron Legacy - "The Grid" playing in my head while solutioning this

[Suman Deb Roy @_RoySD]

Fable is quite amazing, and very surprisingly - token efficient

[Suman Deb Roy @_RoySD]

The game is in less than 24 hrs but with AI, I presume, lawyers can work that fast? Plot twist they run out of tokens

[Brent Scher @BrentScher]

REPORT: Belgium has “lawyered up” to fight back against FIFA’s decision to reinstate Balogun, according to a local reporter for the team.

[Nostalgic Gamer @16bitnostalgia]

Remember in FIFA when you could run away from the ref trying to give you a yellow card? Good times!

[Suman Deb Roy @_RoySD]

You can just do things

[Zealand @theoldzealand]

Yesterday I walked from Times Square to NY/NJ stadium for France v Senegal

I did it completely legally, on sidewalks the entire way, proving once and for all that it is 100% possible to walk to the games in New York

2 states
8 hours
21 miles
43,000 steps

Video coming soon

[Suman Deb Roy @_RoySD]

Wouldn't say impossible, but the attack surface is quite huge so if your tool/paradigm of choice is mech-interp then it appears impossible

[clem 🤗 @ClementDelangue]

Lots of people have known for a while that guardrails for frontier model APIs are very easily jailbroken, quite shallow and impossible to fix. They’re mostly a smokescreen and distraction, in my opinion. We need a different paradigm for AI safety!

[Suman Deb Roy @_RoySD]

This might be the greatest half of football #usmnt has ever played

[Suman Deb Roy @_RoySD]

API next pls

[Gemini Notebook @Gemini_Notebook]

Introducing a more powerful NotebookLM 🚀

Massive upgrades deliver agentic capabilities in chat, more advanced reasoning, and a suite of new output formats. Tackling complex, multi-step research problems has never been easier.

Rolling out now to Google AI Ultra subscribers.

[Suman Deb Roy @_RoySD]

In which some in SV rediscover... "consulting". Almost every attribute here is upheld by traditionally good consultants. And yes, you shouldn't write software from ivory towers that touch millions of people - you have to walk the halls where your users are and breathe the air

[Andrew Ng @AndrewYNg]

One of the new, buzzy jobs in Silicon Valley is the AI Forward Deployed Engineer (FDE), an engineer who is embedded within a client organization to help customize solutions, such as building and tuning agentic workflows that suit the client’s particular needs. I’ve heard from people who are wondering anew about the FDE career path since OpenAI and Anthropic started building new teams to place FDEs within client organizations.

The rise of FDEs for AI workloads is one way AI is creating new jobs (and why the jobpolcalypse narrative of upcoming job market collapse is false -- there will be many AI and non-AI jobs). However, I believe there will be far more AI Engineer jobs than FDEs, as I explain below.

The FDE role was pioneered about two decades ago by Palantir, which sent engineers to government locations to work on secure, air-gapped networks. In addition to having good technical skills, FDEs need communication skills and sometimes business skills. For example, they may need to speak with clients to understand their needs, formulate a strategy to prioritize projects, explain complex technology, and respectfully push back if a client asks for something unrealistic. They’re enjoying a resurgence because of the amount of work involved in taking an off-the-shelf LLM and building it into a custom agentic workflow that fits particular business needs.

However, I believe the number of AI Engineer jobs will be far larger. A company might accept a few FDEs to be embedded within its organization. But most companies will want far more of their own employees working on their projects. While my organizations do hire FDEs, we hire far more AI Engineers! Also, a common client concern is that it is hard to find vendor-neutral FDEs — they are, after all, there to deeply integrate a particular vendor’s product into a company. In this moment when it’s hard to predict which AI service will be the best one in a year’s time, optionality (the ability to pick whatever vendor turns out to fit best in the future) is very valuable. In contrast, letting FDEs tightly bind a company’s processes significantly reduces optionality.

Right now, I see surging demand for AI Engineers who can build software applications using AI software components (like LLM prompting, agentic frameworks, evals, etc.) and effectively use AI coding agents (like Claude Code, Codex, Antigravity CLI, and OpenCode). As the AI Engineer role matures, I expect it to fragment into more specialized roles, like the generic Software Engineer role from decades ago fragmented into frontend, backend, mobile, data engineering, devops, and so on.

What will be the future, specialized AI engineering roles? I don’t know. Perhaps there will be AI FDEs, LLMOps Engineers, Evals Engineers, AI Data Engineers, Harness Engineers, and other roles we don’t have names for yet. But for now, I see a lot of AI engineers who are generalists create a lot of value. Skilled AI Engineers are in very high demand! As our field continues to mature over the coming decade, I look forward to new specializations within AI Engineering that create even more job opportunities.

[Original text: The Batch newsletter]

[Suman Deb Roy @_RoySD]

What a decade. You changed football and the Premier League. Thanks for everything

[Suman Deb Roy @_RoySD]

Waiting for Pep

[Suman Deb Roy @_RoySD]

Although @askalphaxiv is very useful

[Suman Deb Roy @_RoySD]

Same

[Marius Vach @rasmus1610]

Honestly, I love the current trend of publishing research as blog posts instead of arxiv pdfs. @thinkymachines @NoahZiems and @a1zhang come to mind here recently

[Suman Deb Roy @_RoySD]

@CharlotteAlter No, the real tide has turned and academia is beached.

[Suman Deb Roy @_RoySD]

For systems engg. folks: guide them away from failure modes

[François Chollet @fchollet]

A mental model for working with coding agents is that they're blind squirrels running into a maze and bumping into walls. You must place the walls (verifiable constraints) strategically so that they end up in the general region you want them in.

[Suman Deb Roy @_RoySD]

Its still amazing that core GPT is just matrix multiplication stacked in a specific pattern between the token lookup and softmax head at the algorithm level. The rest is predominantly compute allocation

[Suman Deb Roy @_RoySD]

👀

[International Cyber Digest @IntCyberDigest]

❗️🚨 BREAKING: Researchers used Mythos Preview to find the first public macOS kernel memory corruption exploit on Apple's M5 silicon, they give a glimpse into Mythos say it’s really powerful.

Apple spent five years and an estimated several billion dollars building Memory Integrity Enforcement (MIE), the hardware-assisted memory safety system built around ARM's MTE. It was the flagship security feature of the M5 and A19, designed specifically to kill the entire memory corruption bug class.

Researchers from Calif built a working exploit in five days.

According to Apple's own research, MIE disrupts every public exploit chain against modern iOS, including the recently leaked Coruna and Darksword kits. Calif walked into Apple Park this week and handed over the report in person.

Full 55-page technical report drops after Apple patches the vulnerability.

[Suman Deb Roy @_RoySD]

Will an investigative journalist pick this up further https://t.co/edIpKUw9Ao