Phelim Bradley

CEO @Prolific

[Phelim Bradley @Phelimb]

Autonomous agents are good at exploring within a frame, but not at stepping outside it.

Our team ran 50 autoresearch experiments on a DPO task - the agent flagged LoRA by experiment 11 and never followed up. Five minutes of guidance unlocked it 👇🏻

https://t.co/IJyGXedJMH

[Phelim Bradley @Phelimb]

The data quality problem isn't going away anytime soon. About a month ago we announced Prolific's 100% Human Guarantee, but the real work is in the systems that stop threats from getting through in the first place.

I've shared more thoughts here: https://t.co/mjBHyQFF8T

[Business Post @businessposthq]

Prolific, a fast-growing Irish-founded company which has developed an online platform to gather data to train AI models, is set to surpass $500 million (€428 million) in annualised revenue this month. https://t.co/FHLEMRNgeH

[Phelim Bradley @Phelimb]

Paper: https://t.co/NT6QaXDNeA

It's what our sciences team at @Prolific also built HUMAINE to address - LLM evaluation using demographically stratified participants. We're presenting it at #ICLR2026 soon!

https://t.co/G2R7wX06EW

[Phelim Bradley @Phelimb]

Researchers from @Harvard find that LLMs claiming "human-like" performance actually reflect a very specific subset of humanity.

They cluster closest to WEIRD populations (Western, Educated, Industrialized, Rich, Democratic), diverging as psychological distance increases (r ≈ -0.70) 👇🏻

[Phelim Bradley @Phelimb]

@alanchanguk 🚀

[Phelim Bradley @Phelimb]

@SebJohnsonUK @hkanji More of US winning dynamic can be attributed to their institutions and largest enterprises being much more willing to buy from startups than people think

[Phelim Bradley @Phelimb]

AI pollution in human data samples is a hot topic.

Some great work from @andrew_j_gordon et al. showing that concerns here are (generally) overblown, with the majority of platforms empirically showing low levels of AI pollution. https://t.co/6Vir9p9jBE

[Andrew Gordon @andrew_j_gordon]

New preprint out today (https://t.co/UChyCZTcxv). We tested whether AI agents are actually infiltrating online surveys.

Spoiler alert: they aren't

Thread 🧵

[1/9]

[Phelim Bradley @Phelimb]

@jonj @Prolific The default is refunds based on using our bot detection tools. However, if you use independent verification that we can confirm we’ll honour the guarantee.

[Phelim Bradley @Phelimb]

Today we’re launching the Prolific 100% Human Guarantee. If an AI agent is detected in your Prolific study, you'll get twice the cost of that participant response back.

@Prolific’s bot authenticity checks have been doing AI detection work for a while now, tested against major detection methods and achieving 100% accuracy. But it's the years of investment and 50+ checks before participants even see your study that make us especially confident.

Read about our guarantee here, I would love to hear any other feedback: https://t.co/2peGlrzioZ

[Dealroom.co @dealroomco]

Europe has amazing companies, it’s time to celebrate them

Today, with @SebJohnsonUK, we're launching the Scaling Europe Top 50 - the definitive list of Europe's fastest growing companies:

📊 Quantitative, fresh data
❌ No vibes-based ranking, no vanity metrics

We're collecting actual, recorded 2025 revenue data to generate this list, and we've already had over 120 nominations from VCs across Europe. Public nominations are now open (link in thread).

Our goal is simple: To create the most robust ranking of high-growth European tech companies ever produced, and then make as much noise about them as possible. It's time to change the narrative around Europe.

Excited to be working with Scaling Europe @jpmorgan @nebiusai

[Ivan Levingston @IvanLevingston]

scoop: UK chip start-up Fractile seeks to raise $200mn to challenge Nvidia

with A team @tim and @GeorgeNHammond https://t.co/gxi5BdEisT

[Phelim Bradley @Phelimb]

New working paper on online research data quality, led by @univienna reveals that pass rates on quality checks vary wildly by source. Pretty interesting.

Prolific: 90% | Lab: 80% | Bilendi: 73% | Moblab: 55% | MTurk: 9% | AI agents: 0%

https://t.co/ZztDSLoAVn CC @yusof_jeff

[Phelim Bradley @Phelimb]

@HarryStebbings Just waiting for the term sheet 🚀 🇪🇺

[Alex Kendall @alexgkendall]

Day 4, video #4.

This hour-long, zero-shot autonomy video is a route I've done many times: from our London HQ to Heathrow 🇬🇧

Starting in London forced us to build autonomy that could scale. Compared to SF, London has 10x more cyclists, 15x more pedestrians and 20x more roadworks. This is what we've built at @wayve_ai. Check it out!

[Jean P.D. Meijer ― 🇪🇺 eu/acc @initjean]

EU Inc. proposal right now:

- max. 48 hrs and €100 to incorporate
- registration through common EU portal, automatic tax registration
- fully digital process, no notaries
- EU employee stock option plans, taxed only when sold
- simplified insolvency procedures

[Phelim Bradley @Phelimb]

@seanjwestwood We refund any rejected responses, but it's is a fair point that we could go further and be more proactive here. Appreciate the challenge.

[Stefan Schubert @StefanFSchubert]

There are now more tech workers going from the US to Europe than the other way around

[Sovereign AI @UKSovereignAI]

Introducing Sovereign AI, the Government’s new £500m venture fund.

Sovereign AI will support founders from day one to start here, scale here and win everywhere.

https://t.co/frsEPhSDie