Let's walk through hmexx's premises.

Premise 1, that investors overestimate their ability to pick good ideas is undoubtedly true. The best investors are still usually wrong. But the consistent returns that the Sequoias of the world suggest that even if they're usually wrong, they're right enough to make oodles of money.

Premise 2, that an MVP built by a lone founder is just as likely to succeed as a full team on AngelList is almost certainly false. Years of research have demonstrated that founding teams help success. A lone founder can certainly succeed, but is less likely to do so.

Premise 3, that freelancers don't pursue their ideas is absolutely true.

Premise 4, that HackerNews has talented people with good ideas is also absolutely true.

Now comes the key--he's offering $8K (plus the value of his marketing expertise) for 50% of the company.

Let's be generous and assume the marketing consulting is worth $7K. If the average seed investor is paying a $5 million premoney these days, that's a 99.7% discount! Even if a sole founder is 1% as likely to succeed as an AngelList startup, he'll still be 3X ahead.

My conclusion is that hmexx is a marketing genius. He's found a way to buy stock at a 99%+ discount. Also, he's still going to apply a selection process--if Premise 1 were true, he ought to choose at random.

Hmmm, maybe I should invest in him!

Also, just to add to the scrum, I'll double the offer to $10,000!